Page 1
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
A cultivated planet in 2010 – Part 2: The global gridded
agricultural-production maps
Qiangyi Yu
1
, Liangzhi You
1,2
, Ulrike Wood-Sichra
2
, Yating Ru
2
, Alison K. B. Joglekar
3
, Steffen Fritz
4
,
Wei Xiong
5
, Miao Lu
1
, Wenbin Wu
1
, and Peng Yang
1
1
Key Laboratory of Agricultural Remote Sensing (AGRIRS), Ministry of Agriculture and Rural
Affairs/Institute of Agricultural Resources and Regional Planning, Chinese Academy of
Agricultural Sciences, Beijing 100081, China
2
International Food Policy Research Institute (IFPRI), Washington DC, USA
3
GEMS Agroinformatics Initiative, University of Minnesota, Saint Paul, Minnesota, USA
4
International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria
5
International Maize and Wheat Improvement Center (CIMMYT), Texcoco, Mexico
Correspondence:
Wenbin Wu (wuwenbin@caas.cn) and Peng Yang (yangpeng@caas.cn)
Received: 20 January 2020 – Discussion started: 20 March 2020
Revised: 12 October 2020 – Accepted: 21 October 2020 – Published: 21 December 2020
Abstract.
Data on global agricultural production are usually available as statistics at administrative units, which
does not give any diversity and spatial patterns; thus they are less informative for subsequent spatially explicit
agricultural and environmental analyses. In the second part of the two-paper series, we introduce SPAM2010
– the latest global spatially explicit datasets on agricultural production circa 2010 – and elaborate on the im-
provement of the SPAM (Spatial Production Allocation Model) dataset family since 2000. SPAM2010 adds
further methodological and data enhancements to the available crop downscaling modeling, which mainly in-
clude the update of base year, the extension of crop list, and the expansion of subnational administrative-unit
coverage. Specifically, it not only applies the latest global synergy cropland layer (see Lu et al., submitted to
the current journal) and other relevant data but also expands the estimates of crop area, yield, and production
from 20 to 42 major crops under four farming systems across a global 5 arcmin grid. All the SPAM maps are
freely available at the MapSPAM website (http://mapspam.info/, last access: 11 December 2020), which not
only acts as a tool for validating and improving the performance of the SPAM maps by collecting feedback
from users but is also a platform providing archived global agricultural-production maps for better targeting the
Sustainable Development Goals. In particular, SPAM2010 can be downloaded via an open-data repository (DOI:
https://doi.org/10.7910/DVN/PRFF8V; IFPRI, 2019).
1
Introduction
Civilization is founded on the agricultural use of land (Fu
and Liu, 2019), which remains as important today as it was
10 000 years ago (Lev-Yadun et al., 2000). Agricultural land,
which refers to the land area that is arable, under permanent
crops, and under permanent meadows and pastures accord-
ing to the Food and Agriculture Organization of United Na-
tions (FAO), is currently 4.9 billion ha in 2019. This is 37.6 %
of the earth’s terrestrial surface – the largest use of land on
the planet. Historically, the agricultural use of land has trans-
formed ecosystem patterns and processes across most of the
terrestrial biosphere (Ellis et al., 2013). The way we use agri-
cultural land will significantly determine whether we are able
to solve the multiple challenges embodied in the 17 Sustain-
able Development Goals (SDGs), e.g., feeding the world’s
growing population, mitigating climate change, and halting
biodiversity loss (FAO, 2018; Ehrensperger et al., 2019). As
the fundamental connection between people and the planet,
the spatiotemporal characteristics of agricultural land is im-
Published by Copernicus Publications.
Page 2
3546
Q. Yu et al.: A cultivated planet in 2010 – Part 2
portant for the anthroposphere and beyond as such informa-
tion allows us to undertake more responsive and evidence-
based analysis on the interaction and better resource alloca-
tion across land, water, energy, and the environment.
Cropland mapping has made great progress in the past few
decades and provided great support for global agricultural
monitoring and assessment. For example, it allows us to be
able to know where agriculture has infringed into natural
ecosystems and where cropland has been taken as a conse-
quence of urbanization (Chen et al., 2015; Gong et al., 2019).
However, this type of work mainly focuses on the agricultural
changes at the land cover level, without paying attention to
the subtle characteristics at the land use and land manage-
ment level (Verburg et al., 2011). These subtle level charac-
teristics related to agricultural production, ranging from crop
allocation to land use intensity, are the core of agricultural
management and have been proven to have equally impor-
tant impacts on food systems (Sun et al., 2018; Pretty, 2018),
climate systems (Searchinger et al., 2018; Bonan and Doney,
2018), and ecosystems (Peters et al., 2019; Poore and Ne-
mecek, 2018). Yet data on global agricultural production are
usually representative at national and subnational adminis-
trative units (e.g., provinces, districts). This level of statistics
does not give a sense of the diversity and spatial patterns in
agricultural production and is not spatially explicit, which is
critical for many environmental and ecological assessments
(Yu et al., 2012).
There are a few attempts to develop global spatially ex-
plicit datasets on agricultural production by fusing census
statistics with maps of agricultural land cover (Fig. 1). Leff
et al. (2004) applied a simplified proportional disaggregat-
ing approach and mapped the global harvested area of 18
major crops circa 1992. By using a similar approach, Mon-
freda et al. (2008) mapped both harvested area and yield for a
full coverage of 175 crops circa 2000 (the dataset is referred
to as M3 hereafter). Portmann et al. (2010) developed the
MIRCA (Monthly Irrigated and Rainfed Crop Areas) dataset
that contains the harvested area for 26 crops circa 2000 by
using M3 as a starting point. It further allocates the total har-
vested area for each crop into rainfed and irrigated areas.
Fischer et al. (2012) developed the GAEZ dataset (Global
Agro-ecological Zones), which contains the potential har-
vested area and yield for 23 crops circa 2000 considering
the crop-specific agroclimatic and edaphic suitability crite-
ria. You and Wood (2006) developed the Spatial Production
Allocation Model (SPAM) firstly at the continental scale then
subsequently at the global scale by using an entropy-based
model to downscale crop production. The first global SPAM
dataset is available for the year 2000, at the time when M3,
MIRCA, and GAEZ were also available (Fig. 1).
Changes in agricultural lands over time are as important
as over space, especially given that the changes in cropping
pattern and crop yields are more frequent than those at the
land cover level (Verburg et al., 2011). While there are four
spatially explicit datasets on global agricultural production
available around the year 2000 (Anderson et al., 2015), three
of them, i.e., M3, MIRCA, and GAEZ, are no longer avail-
able after 2000. Agricultural-production systems are con-
stantly changing, and these changes are not trivial. How-
ever, a lot of recent agricultural and environmental assess-
ments were still based on those maps produced decades ago
(Deutsch et al., 2018; Nanni et al., 2019; Estes et al., 2018;
Prestele et al., 2018; Erb et al., 2018; Porwollik et al., 2019;
Yu et al., 2017b), suggesting that an update of existing global
agricultural-production maps is very desirable for subsequent
analysis.
SPAM had committed to updating maps every 5 years
(You et al., 2014; Wood-Sichra et al., 2016), which sub-
stantially fills the data gap and extends the work for global
agricultural-production mapping by operating a global grid-
scape at the confluence between earth and farming systems
in multiple time stages. The SPAM model has become a crit-
ical tool to many initiatives within and beyond the Consulta-
tive Group for International Agricultural Research (CGIAR).
Moreover, SPAM data are frequently downloaded and widely
used by researchers and analysts from international origi-
nations, academia, and governments agencies all over the
world. The global spatially explicit datasets in multiple time
stages enable scientists as well as policymakers to better ad-
dress the global change challenges within the anthroposphere
and beyond, such as targeting agricultural and rural devel-
opment policies and investments and increasing food secu-
rity and growth with minimal environmental impacts. Suc-
cessful examples include AGRODEP (African Growth and
Development Policy Modeling Consortium) Library (http:
//www.agrodep.org/fr/node/1794, last access: 11 December
2020), GEOGLAM (Global Agricultural Monitoring Ini-
tiative) (http://www.geoglam.org, last access: 11 Decem-
ber 2020), USAID (United States Agency for International
Development) Feed the Future Innovation Lab for Small-
scale Irrigation (https://ilssi.tamu.edu/, last access: 11 De-
cember 2020), Africa Infrastructure (https://openknowledge.
worldbank.org/handle/10986/2692, last access: 11 December
2020), and so on. In this paper, we introduce SPAM2010, the
latest update of the SPAM family. The next section gives an
overview of the SPAM model. Section 3 provides a detailed
description and improvements of SPAM2010. Section 4 in-
troduces the data preparation, and Sect. 5 presents some of
the results produced by SPAM2010. Finally, we conclude
with some advice on using the maps and our own plan for
the future of SPAM.
2
SPAM overview
The main purpose of SPAM is to disaggregate crop statistics
(e.g., harvested area, production quantity, and yield) by dif-
ferent farming systems and to further allocate such disaggre-
gated statistics into spatially gridded units (Fig. 2). In SPAM,
disaggregation is processed before allocation because crop
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 3
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3547
Figure 1.
Overview of the global spatially explicit datasets on agricultural production. Each dataset is plotted in a coordinate system with
the
x
axis representing the time span and the
y
axis representing the number of crops that have been included. For each dataset, the first row
indicates the major measurement(s) of agricultural production, the second row indicates the cropland cover layer, and the third row indicates
the main approach for allocating production. The dashed line within the chart indicates the evolution of a dataset family.
yields are likely to be substantially different between differ-
ent farming systems (e.g., irrigation versus rainfed) even at
the same location. The whole procedure entails a data fusion
approach that combines information from different sources
and at different spatial scales by deploying various matching
and calibration processes. Then all the data elements are pro-
cessed by the optimization model, which generates results at
the grid level (Fig. 2).
The SPAM methodology was first developed in a trial
project for six major crops in Latin America and the
Caribbean by combining satellite imagery and crop statis-
tics. Later on, it was used to derive regional estimates of
spatially disaggregated crop production in Brazil and sub-
Saharan Africa (You and Wood, 2006; You et al., 2009).
Over the years the model has evolved, adding more crops
and using additional data and increasingly complicated opti-
mization equations as well as expanding to global coverage
(You et al., 2014). The SPAM methodology is different from
its counterparts. For example, M3 has no distinction across
farming systems and is allocated proportionally (within each
crop) to each grid cell within each subnational unit; hence the
M3 dataset provides interpolated estimates of output by crop
at the resolution of the satellite data (Fig. 1). SPAM not only
considers the crop yield variation across farming systems but
also assigns production weighted by price to grid cells rather
than pure proportionality (Donaldson and Storeygard, 2016).
Moreover, MIRCA and GAEZ focus more on the biophysical
aspects of agricultural production, while SPAM uses a trian-
gulation of any and all relevant background and partial infor-
mation, which includes not only national or subnational crop
production statistics and satellite data on land cover but also
maps of irrigated areas, crop potential suitability, secondary
data on population density, market accessibility, cropping in-
tensity, and crop prices (Fig. 2).
The SPAM model produces global gridded maps of agri-
cultural production at a 5 arcmin spatial resolution. The first
SPAM maps, known as SPAM2000, represent global agricul-
tural production circa 2000 for 20 crops, with the exception
of a few small island states and conflict zones (You et al.,
2014). Subsequently, the SPAM maps have been updated ev-
ery 5 years. SPAM2005 acts as an intermediate update which
expands the coverage of crops from SPAM2000. The 42 crop
categories are further adopted in SPAM2010 (Fig. 1).
3
Model improvement for SPAM2010
There are three submodules in a standardized SPAM model:
disaggregation, optimization, and allocation. We conceptu-
alize the SPAM2010 based on this general setting, while it
adds further methodological and data improvements, which
mainly include the update of base year, the expansion of sub-
national administrative-unit coverage, the extension of crop
list, and the substitution of the latest hybrid cropland map
as the basic allocation layer. Considering the huge number
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 4
3548
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Figure 2.
The overall structure of the SPAM model. The rhombuses indicate spatial data inputs and outputs, while the other shapes indicate
nonspatial data inputs (see the detailed data description in the following section). The orange color indicates how crop statistics are disaggre-
gated by administrative unit (
k
), crop type (
j
), and farming system (
l
). The green color indicates how the spatial parameters are collected and
prepared at a unified spatial resolution (
i
) and in a harmonized manner. The yellow color indicates the spatial allocation inputs/outputs. The
darker colors, either in orange or in green, highlight the essential elements in SPAM: the former indicates the farming system disaggregation
scheme, while the latter indicates (i.e., priors of physical area) a key parameter with which the spatial and nonspatial data are connected and
the iterative spatial allocation is able to take place.
of input data and multiple-year efforts, such an update is not
trivial and will be critical for the user community. In this sec-
tion, we briefly introduce the model structure and how these
submodules are processed and connected.
3.1
Disaggregation
The first step for SPAM is to disaggregate crop statistics of
agricultural production (e.g., the yield, harvested area, and
total production) by administrative-unit levels (
k
), crop type
(
j
), and farming system (
l
) from coarser scale to finer scale
(illustrated by orange shapes in Fig. 2). For example, the
national-level statistics are disaggregated into subnational
levels, statistics for crop aggregates are divided into individ-
ual crop types, and the crop statistics are further separated
by rainfed and irrigated conditions. Disaggregation is a non-
spatial module. For the administrative unit (ADM), we con-
sider three levels –
k
=
0 (national level), 1 (subnational level
1), and 2 (subnational level 2) – and refer to the country-
specific administrative level as the statistical reporting units
(SRUs; SRU
=
k
0
,
k
1
or
k
2
). In general, the SPAM model
will have better performance if crop statistics are more dis-
aggregated by the ADM. Therefore, we prefer to collect crop
statistics for ADM1 and ADM2, despite statistics mostly be-
ing available at the ADM0 level and the subnational cover-
age always being less complete. Comparing to the previous
SPAM products, the subnational coverage percentage has in-
creased markedly for SPAM2010, which is described in de-
tail in Sect. 4.1.
We improve the model capacity in SPAM2010 as well: we
simultaneously allocate 42 crops and crop aggregates (versus
the 20 crops and crop aggregates in SPAM2000) and consider
four farming systems for each crop (Fig. 2). In SPAM2010,
we keep the farming systems conceptualized for SPAM2000,
which have proven to be useful to represent the different crop
performances under different management systems; e.g., the
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 5
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3549
irrigated yields of a particular crop are likely to be substan-
tially different from the corresponding rainfed yields. The
four farming systems are defined as follows:
–
The irrigated farming system (I) refers to the crop area
equipped with either full or partial control irrigation.
Normally the crop production on the irrigated fields uses
a high level of inputs such as modern varieties and fer-
tilizer as well as advanced management such as soil and
water conservation measures.
–
The rainfed high-input farming system (H) refers to the
market-oriented crop area, which uses high-yield vari-
eties; machinery with low labor intensity; and optimum
applications of nutrients and chemical pest, disease, and
weed control.
–
The rainfed low-input farming system (L) refers to crop
area which uses traditional varieties and mainly man-
ual labor without (or with little) application of nutrients
or chemicals for pest and disease control. Production is
mostly for personal consumption.
–
The rainfed subsistence farming system (S) is intro-
duced to account for situations where cropland and suit-
able areas do not exist, but farmland is still present in
some way. Production is mostly for personal consump-
tion, which is also low-input.
The four conceptualized farming systems are mainly delin-
eated by the water supply system and inputs used by farmers,
despite global data on farming system shares for each crop
being largely absent. For a small number of large countries,
e.g., Brazil, China, India, Russia, the United States (see more
details in Sect. 4.1), we have data on farming system shares
at the ADM1 level. For the other countries we first assign the
national farming system shares to each ADM1 level and then
adjust individual ADM1 farming system shares in light of the
supporting evidence. For example, if the national share for ir-
rigation of wheat was 30 %, we assign that to all ADM1 units.
Then we look at individual units, and if supporting evidence
(e.g., the Global Map of Irrigation Areas, GMIA, data) indi-
cates that there was no irrigated area present in a particular
AMD1 unit, we set the irrigation share of wheat to 0 in that
administrative unit. Finally the farming system shares at the
national level are recalculated as the weighted average of the
adjusted ADM1 estimates. For a few countries which have
very limited data accessibility, experts may give their opin-
ions. For example, it was often necessary to use farming sys-
tem shares from one crop as proxies for similar crops (e.g.,
farming system shares for beans are used for all pulses) or to
apply shares from one country to similar countries (e.g., the
geographically smaller countries in the Middle East, includ-
ing Kuwait, Oman, and Qatar, are assigned the same farming
system shares).
For irrigated farming systems, the crop-specific shares are
derived by dividing the harvested area cultivated under full-
control irrigation obtained from AQUASTAT, MIRCA, and
country-level statistics by the overall harvested area. For
rainfed farming systems, crop-specific shares are primarily
estimated based on generalized assumptions for individual
countries and crops. For example, all cereals in western Eu-
rope are produced with high inputs, whereas 20 % of cereals
in sub-Saharan Africa are grown under a subsistence farm-
ing system. We also assume fertilization as a proxy for high-
input use, so if irrigated crop areas and overall fertilized and
nonfertilized areas of a crop are known, it is possible to de-
duce rainfed–high shares by subtracting the irrigated areas
from fertilized areas. The remainder of fertilized area will
be then classified as rainfed–high, and the nonfertilized ar-
eas will be further split between rainfed–low and rainfed–
subsistence. In addition, the shares of rainfed–subsistence are
assigned when there is not enough suitable area for rainfed–
low conditions to satisfy the completeness of disaggregated
crop statistics in terms of area extent and/or production quan-
tity. In such cases a portion of the rainfed–low statistics were
assumed to stem from rainfed–subsistence. Although disag-
gregation is a nonspatial module of SPAM, it is applied inter-
actively with the spatial modules by the support of multiple
spatial data and nonspatial data, which are elaborated in de-
tail in the section “Data preparation for SPAM2010”.
3.2
Optimization
The core part of the SPAM model is the cross-entropy mod-
ule (illustrated by the dashed green frame in Fig. 2), which
is used to achieve the allocation for each spatial grid (
i
).
It works by (iteratively) minimizing the error between the
preallocated shares of physical area (
π
ij l
) and the allocated
shares of physical area (
s
ij l
) in each pixel
i
by crop
j
and
production system
l
:
min
{
s
ij l
}
CE
s
ij l
, π
ij l
=
X
i
X
j
X
l
s
ij l
ln
s
ij l
−
X
i
X
j
X
l
s
ij l
ln
π
ij l
,
(1)
where CE is the abbreviation for cross-entropy, which is de-
fined as the log function of probability. The difference be-
tween
{
s
ln
s
}
versus
{
s
ln
π
}
means the estimated probability
s
and its prior probability
π
are minimized subject to certain
constraints.
i. Constraint specifying the range of allocated physical-
area shares:
0
≤
s
ij l
≤
1
,
∀
i
∀
j
∀
l
;
(2)
ii. Constraint specifying the sum of allocated physical-area
shares within a grid:
X
i
s
ij l
=
1
,
∀
j
∀
l
;
(3)
iii. Constraint specifying that the sum of allocated physical
area over all crops and farming systems within a grid
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 6
3550
Q. Yu et al.: A cultivated planet in 2010 – Part 2
should not exceed the actual cropland within the same
grid:
X
j
X
l
AdjCropA
j lk
SRU
×
s
ij l
≤
AdjCropLand
i
,
∀
i
∈
k
SRU
;
(4)
iv. Constraint specifying that the allocated physical area by
grid, crop, and farming system should not exceed the
suitable area within the grid with the corresponding crop
and farming system:
AdjCropA
j lk
SRU
×
s
ij l
≤
AdjSuitArea
ij l
,
∀
i
∈
k
SRU
;
(5)
v. Constraint specifying that the sum of allocated physi-
cal area over all farming systems within a subnational
unit should be equal to the sum of statistical physical
area over all farming systems within the corresponding
subnational unit:
X
i
∈
k
SRU
X
l
AdjCropA
j lk
SRU
×
s
ij l
=
X
l
AdjCropA
j lk
SRU
,
∀
j
∈
P ,
(6)
where
P
is the set of commodities for which subnational
statistics exist;
vi. Constraint specifying that the sum of allocated physical
area under an irrigated farming system within the grid
should not exceed the area equipped for irrigation in the
grid:
X
j
∈
Q
AdjCropA
j I k
SRU
×
s
ij
I
≤
AdjIrrArea
i
,
∀
i
∈
k
SRU
,
(7)
where
Q
is the set of commodities which are fully or
partly irrigated within grid
i
.
Shares
s
ij l
are the probability values between 0 and 1:
s
ij l
=
AllocA
ij l
AdjCropA
j l
,
(8)
where AdjCropA
j l
is the total physical area of a given SRU
for crop
j
at input level
l
to be allocated. AllocA
ij l
is the area
allocated to grid
i
for crop
j
at input level
l
.
π
ij l
indicates the decision to produce a particular crop
under a specific production system, which is normally de-
pendent on both biological and economic factors. However,
subsistence farmers mainly grow crops for their own con-
sumption, largely uncoupled from price, market access, or
crop-potential-suitability conditions. Therefore, we first as-
sume that the prior allocation for subsistence physical area
(CropA
ij
S
) in grid
i
by crop
j
under this circumstance is
simply dependent on rural population density:
CropA
ij
S
=
AdjCropA
j k
S
×
AggRurPop
i
P
i
∈
k
AggRurPop
i
∀
i
∀
j,
(9)
where AdjCropA
j k
S
is the generated physical area for crop
j
at the subsistence farming system for the given SRU
k
, and
AggRurPop
i
is the rural population density at grid
i
(see the
detailed description in Sect. 4).
Then for the three remaining farming systems, we assume
that the potential unit revenue of planting a certain crop
(Rev
ij l
) would affect farmers’ crop choices:
Rev
ij l
=
AdjCropA
j lk
SRU
P
i
AdjCropland
i
×
Price
j
×
Access
i
×
PotYield
ij l
,
(10)
where AdjCropland
i
, Price
j
, Access
ij
, and PotYield
ij l
are
the adjusted cropland area, market price, accessibility pa-
rameter, and potential-yield values for crop
j
in farming sys-
tem
l
and grid
i
(see the detailed description in the following
Sect. 4).
Then we assume that the priors for the remaining three
farming systems are mainly influenced by the estimated rev-
enue, cropland area, and irrigated area.
For an irrigated farming system (I),
CropA
ij
I
=
AdjIrrArea
i
×
Rev
ij
I
P
j
Rev
ij
I
∀
i
∀
j.
(11)
For rainfed–high (H) and/or a rainfed–low (L) farming sys-
tems,
CropA
ij l
=
AdjCropLand
i
−
AdjIrrArea
i
−
CropA
ij
S
×
Rev
ij l
P
j
Rev
ij l
, l
=
HL
∀
i
∀
j,
(12)
where AdjCropLand
i
and AdjIrrArea
i
are the cropland area
and irrigated area at grid
i
(see the detailed description in the
following Sect. 4).
Finally, the main inputs for the optimization procedure are
converted to shares and written as
π
ij l
=
CropA
ij l
P
i
∈
k
SRU
CropA
ij l
.
(13)
The optimization module in SPAM2010 is almost the same
as that in previous versions. We apply the cross-entropy pro-
cess in the General Algebraic Modeling System (GAMS),
which ensures that the optimization procedure iterates un-
til a solution is found. Once the allocation is successful,
meaning that an optimal or locally optimal solution has been
found, the routine immediately returns the allocated physi-
cal area (AllocA
ij l
) by grid
i
, crop
j
, and farming system
l
, and the program continues with postprocessing automati-
cally (Fig. 2). If the solution is infeasible or nonoptimal, the
program stops, allowing for manual scrutiny, adjustment, and
rerun (see data harmonization in the following section).
3.3
Allocation
Using the results of the optimization, the allocation mod-
ule produce maps of harvested area (AllocH
ij l
), yield
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 7
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3551
(AllocY
ij l
), and production quantity (AllocP
ij l
) for each grid
i
by crop
j
and farming system
l
(Fig. 2). For harvested area,
we convert the allocated physical area (AllocA
ij l
) to allo-
cated harvested area (AllocH
ij l
) by multiplying by cropping
intensity (CropIntensity
j lk
):
AllocH
ij l
=
AllocA
ij l
×
CropIntensity
j lk
.
(14)
For yield, we first calculate an average potential yield
(PotYield
j lk
SRU
) within an SRU using the allocated harvested
area as weight, then the allocated yield (AllocY
ij l
) is esti-
mated as
AllocY
ij l
=
AdjPotYield
ij l
×
AdjCropY
j lk
SRU
PotYield
j lk
SRU
,
(15)
where the average potential yield is calculated as
PotYield
j lk
SRU
=
P
i
∈
SRU
AdjPotYield
ij l
×
AllocH
ij l
P
i
∈
k
SRU
AllocH
ij l
.
(16)
We finally estimate the production quantity (AllocP
ij l
) as
AllocP
ij l
=
AllocH
ij l
×
AllocY
ij l
.
(17)
4
Data preparation for SPAM2010
The largest amount of effort to create a SPAM map is spent
on identifying, collecting, and harmonizing data. For the pro-
duction of SPAM2010, we collect raw data from two major
sources: we first collect nonspatial crop statistics for the data
disaggregation process; we then collect and/or create mul-
tiple spatially explicit constraint maps at a 5 arcmin resolu-
tion from both biophysical and socioeconomic aspects for the
spatial optimization and allocation processes. Afterwards, we
introduce how these multisourced data are harmonized and
how data adjustment is taking place.
4.1
Crop statistics
4.1.1
Crop statistics disaggregated by administrative
units
We start with the administrative units (
k
) for which we
have been able to obtain crop production statistics (Fig. 2).
We primarily used the FAO’s Global Administrative Unit
Layers (GAUL) at both the national and subnational levels
to relate the tabulated crop statistics to gridded data dur-
ing the allocation process. GAUL contains shapefiles for
three administrative-level units: ADM0 (national level 0),
ADM1 (subnational level 1), and ADM2 (subnational level
2). Shapefiles from the Database of Global Administrative
Areas (GADM) are used for ADM1 and ADM2 in China
since they proved to be easier to match to the statistics.
We collect crop statistics from FAOSTAT, EUROSTAT,
CountrySTAT, ReSAKSS, national statistical offices, min-
istries of agriculture or planning bureaus of individual coun-
tries, household surveys, and a variety of ad hoc reports re-
lated to a particular crop within a particular country (Fig. 2).
SPAM estimates are most dependent on the degree of dis-
aggregation of the underlying national and subnational pro-
duction statistics, so it is important to identify and collect as
many subnational statistics as possible (Joglekar et al., 2019).
Although we prefer to collect crop statistics for ADM1 and
ADM2 and run the model at the ADM1 level for all coun-
tries, crop statistics are unfortunately mostly available at the
ADM0 level, the subnational coverage being less complete.
Therefore, for most countries we run SPAM at an ADM0
level, except for some (geographically) large countries that
are modeled at an ADM1 level. We summarize the subna-
tional data coverage by region in Table 1. We present the
detailed procedure for collecting crop statistics in the Sup-
plement (Sect. S1), which further contains a table listing all
countries that are modeled at an ADM1 level (Table S1) and
a table listing the sources of crop statistics by country and
subnational coverage (Table S2) for all countries.
We collect data in all the ADM1 units in the United States,
Russia, and Canada and at least 80 % of the ADM1 units
for the remaining regions worldwide, while Europe, the Mid-
dle East, Oceania, Russia, and sub-Saharan Africa have data
collected on the full set of crops in below 80 % of their
ADM2 units. This coverage is substantially improved for
SPAM2010 in comparison to that for SPAM2005, which is
only 66.2 % and 43.2 % for ADM1 and ADM2, respectively
(Table 1).
Monfreda et al. (2008) reported that 81 % of the global
harvested-area data in their M3 in the year 2000 came from
subnational sources, but they do not distinguish coverage by
subnational levels 1 and 2. SPAM often has higher levels
of subnational coverage than M3, especially in Africa and
the former states of the Soviet Union. This can be seen in
SPAM2005; e.g., 93.4 % of global data came from ADM1
sources and 54.6 % from ADM2 sources (Wood-Sichra et
al., 2016). In SPAM2010, such coverage rates are further in-
creased to 96.1 % and 68.0 %, respectively (Table 1).
4.1.2
Crop statistics disaggregated by crop types
We simultaneously allocate 42 crops and crop aggregates
(
j
) for SPAM2010 (Fig. 2). The crop categories are driven
by the definitions of the FAO’s Statistical Database (FAO-
STAT). Comprised of 33 individual crops (e.g., wheat, rice,
maize, barley, potato, bean, cotton) and nine crop aggregates
(e.g., other cereal, vegetables), the SPAM2010 crop list cov-
ers all crops reported by the FAO, except for explicit fod-
der crops (mostly grasses), which are not modeled. When
multiple FAO crops fall into a single SPAM2010 crop cat-
egory (e.g., vegetables), the FAO’s corresponding area and
production data were summed up, and yields were calculated
as a weighted average. We present the detailed procedure for
aligning the crop types in the Supplement (Sect. S2), which
further contains a full list of crops and their respective FAO
code (Table S3).
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 8
3552
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Table 1.
Subnational coverage of crop production statistics by region.
Full crop coverage
National harvested-area coverage
Region
Countries
ADM1
ADM2
ADM1
ADM2
Harvested area
ADM1
ADM2
SPAM2010
(Count)
(Percent)
(1000 ha)
(Percent)
Asia
25
477
9513
90.9
86.4
535 759
99.0
84.3
Canada
1
13
202
100.0
81.3
25 841
100.0
85.0
Europe
47
581
3323
86.0
74.0
169 506
91.6
45.3
Central America
42
479
11 721
84.7
92.7
146 396
97.5
81.0
Middle East
14
167
949
81.1
53.6
23 220
79.2
3.4
Northern Africa
6
132
1685
80.6
85.1
21 345
84.0
58.2
Oceania
19
68
789
85.5
48.8
26 379
97.3
1.5
Russia
1
91
91
100.0
0.0
52 079
100.0
100.0
Sub-Saharan Africa
50
687
3928
79.0
68.6
215 896
91.5
30.9
United States
1
51
3106
100.0
87.0
98 991
99.8
94.0
Total
206
2746
35 307
85.1
85.2
1 315 412
96.1
68.0
SPAM2005
Total
201
2799
33 425
66.2
43.2
1 239 026
93.4
54.6
Source
: assembled by authors. Note: full crop coverage refers to the percentage of crops at the administrative level with positive values or 0 in relation to all possible crops. Percent
of national harvested area covered by ADM1 or ADM2 is the share of national area harvested reported by ADM1 or ADM2 units. In Russia we had no data for ADM2 units. The
last row presents an overview of data coverage applied for SPAM2005 as a comparison.
We collect statistics on harvested area (H), production (P),
and yield (Y; CropHPY) by each crop
j
in each administra-
tive unit
k
for data disaggregation (Fig. 2). We prepare data
for the model based on the 2009–2011 average of the crop
production statistics (AvgCropHPY
j k
). If data are missing
from this time period, we use the average from the avail-
able data spanning the closest years between 2005 and 2015.
We make corrections for discrepancies in statistical report-
ing units, crop names, and units of measurement during the
initial cleaning phase of the data. For example, we adjust
all national and subnational statistics (AdjCropHPY
j k
) us-
ing the national 2009–2011 average from FAO. In order to
improve the comparability of the crop production statistics
across countries, we explicitly distinguish between crops not
grown in an area (coded as 0) and crop data that are not avail-
able for an area (coded as a missing value). Despite the possi-
ble uncertainties in FAO data, it has been chosen as the base-
line in the adjustment of country statistics mainly because
(1) FAO data are the most widely acknowledged global agri-
cultural statistics and are hence the most appropriate source
for the purpose, and (2) SPAM products have been used
by many global models such as IMPACT from the Interna-
tional Food Policy Research Institute (IFPRI; https://www.
ifpri.org/project/ifpri-impact-model, last access: 11 Decem-
ber 2020) and GLOBIOM from the International Institute
for Applied Systems Analysis; https://iiasa.ac.at/web/home/
research/GLOBIOM/GLOBIOM.html, last access: 11 De-
cember 2020). These models use FAO country data for cross-
country comparisons, and they need our maps to be con-
sistent with FAO data. In fact, the idea of conceptualizing
SPAM is to spatially allocate statistics from administrative
units to spatial grids, and the maps could be easily adjusted
to any other country data. We present the detailed procedure
for adjusting the crop statistics in the Supplement (Sect. S3).
4.1.3
Crop statistics disaggregated by farming systems
We elaborated the disaggregation module for obtaining the
farming system shares by crop
j
and administrative unit
k
(Percent
j lk
) in Sect. 3.1. In some countries there are statis-
tics, in others experts may give their opinions, or assumptions
are made as to how some crops are grown in a similar way
as other crops. Supplementing Sect. 3.1, we present more
details on the procedure for obtaining the farming system
shares in the Supplement (Sect. S4), which further contains a
table listing the sources of subnational farming systems data
(Table S4) and a table listing the farming system shares by
crop groups and selected countries (Table S5). For example,
shares of the
irrigated farming system
were taken directly for
country statistics like Brazil, China, and the United States
at ADM1. For some countries these figures were found in
MIRCA, and yet for the rest of the countries, AQUASTAT
provides information on irrigated areas per crop at the na-
tional level. We are able to source data on farming system
shares at the ADM1 level for limited large countries (Ta-
ble S4). Based on this list we showcase the shares of pro-
duction under irrigated and rainfed systems for selected crop
groups and countries (Table S5). We choose Brazil, China,
Ethiopia, France, India, Indonesia, Nigeria, Turkey, and the
United States because they vary in agroecology, region, in-
come level, and geographical size. For cereal crops, the three
Asian countries (China, India, and Indonesia) have the high-
est shares of irrigated area, whereas the two sub-Saharan
countries (Ethiopia and Nigeria) have the lowest shares of ir-
rigated area. For root, tuber, and pulse production, the United
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 9
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3553
States and both European countries have the highest shares of
irrigated areas, while the sub-Saharan countries again have
less than 1 % each. Aggregating across all crops, the three
Asian countries rank highest in terms of irrigated area shares,
while the two sub-Saharan countries rank lowest.
We disaggregate the adjusted statistics on harvested area
and yield (AdjCropHPY
j k
) for each of the four farm-
ing systems (Fig. 2). Harvested area by farming system
l
(AdjCropH
j lk
) is directly calculated by multiplying the
farming system shares (Percent
j lk
), while the yields by farm-
ing system
l
(AdjCropY
j lk
) are more complicated to calcu-
late. Here we consider not only the farming system shares
but also the yield conversion factors (determined by expert
judgment) to distinguish the yield variations for irrigated
versus rainfed systems and rainfed–high versus rainfed–low
systems. We present the detailed procedure for disaggregat-
ing the crop statistics by farming systems in the Supplement
(Sect. S5), which further contains a list of the yield conver-
sion factors, i.e., both the factor of crop yield under irrigated
versus crop yield under rainfed (with an “I”) and that of yield
under rainfed high-input versus yield under rainfed low-input
(with an “R”), for selected crops and countries (Table S6).
4.1.4
Physical area
We create a new variable – physical area (AdjCropA, i.e.,
the area footprint of the crop irrespective of the number of
times per year the same area was planted and harvested) – for
the model, recognizing that crop production may take place
over several seasons within a year. SPAM does not have a
direct mechanism for modeling sequential or intercropping
processes, and thus we use harvested area and cropping in-
tensity (CropIntensity) per crop as a proxy for these pro-
cesses:
AdjCropA
j lk
=
AdjCropH
j lk
CropIntensity
j lk
,
(18)
where AdjCropA
j lk
indicates the generated physical area by
crop
j
, farming system
l
, and administrative unit
k
.
Implementing the crop allocation calculations by farming
system enables more flexibility when accounting for vari-
ation in these cropping intensity practices. However, such
data are still scarce. Only some country statistics have such
figures, e.g., Bangladesh and India; thus we rely primarily
on expert judgment to seek information on the number of
cropping seasons by crop, farming system, and country. We
present the detailed procedure for generating physical area
in the Supplement (Sect. S6), which further contains a table
listing CropIntensity
j lk
by crop groups and selected coun-
tries (Table S7).
4.2
Spatial constraints
4.2.1
Cropland extent
We apply an already-classified land cover image – where
cropland has been identified (CropLand) – to determine the
places where production statistics can be allocated. Com-
paring to SPAM2000 and SPAM 2005, SPAM2010 updates
not only the statistics but also the cropland distribution: it
uses the global cropland synergy map with a spatial resolu-
tion of 500 m circa 2010, jointly produced by CAAS (Chi-
nese Academy of Agriculture Sciences) and IFPRI (Fig. 1).
The CAAS–IFPRI cropland dataset fuses national and subna-
tional statistics with multiple existing global land cover maps
including GlobeLand30, CCI-LC, GlobCover 2009, MODIS
C5, and Unified Cropland. It reports three major parameters
by grid around the year 2010: the median and maximum
cropland percentage (MedCropLand
i
and MaxCropLand
i
)
and a confidence score between 0 and 1 in the cropland
estimation (ProbCropLand
i
). Although the synergy dataset
does not delineate the geography of specific crops, it desig-
nates the total cropland extent with a higher accuracy than
the input datasets and tries to be consistent with admin-
istrative cropland statistics. The detailed description of the
CAAS–IFPRI cropland dataset is submitted as a parallel pa-
per (see Lu et al., 2020). Before using the cropland extent
in SPAM2010, we aggregate the cropland synergy map from
500 m grid cells to 5 arcmin grid cells for the three major
parameters. We present the cropland data preparation in the
Supplement (Sect. S7), which further contains the resam-
pled maps on median cropland (AggMedCropLand
i
), max-
imum cropland (AggMaxCropLand
i
), and cropland confi-
dence (AggProbCropLand
i
; Fig. S1).
4.2.2
Crop potential suitability
We estimate the crop suitable area (SuitArea) from
GAEZv3.0 to consider the spatially varied potential suitabil-
ity for different crops in terms of different thermal, moisture,
and soil requirements as an allocating parameter. GAEZv3.0
produces a 5 arcmin gridded suitability index for 49 ma-
jor crops, four input levels (i.e., high, intermediate, low, or
mixed), and two main water regimes (i.e., irrigated or rain-
fed). The major crops surveyed by GAEZ include most of
the SPAM2010 crops; those not included are assigned values
from similar GAEZ crops. We present the detailed procedure
for estimating the suitable area (SuitArea
ij l
) for grid
i
, crop
j
, and input
l
in the Supplement (Sect. S8), which further
contains a table illustrating the concordance between GAEZ
crops and SPAM2010 crops (Table S8) as well as maps of
suitable areas for maize irrigated rainfed–high and rainfed–
low farming systems (Fig. S2).
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 10
3554
Q. Yu et al.: A cultivated planet in 2010 – Part 2
4.2.3
Irrigated area
We adopt the irrigated area (IrrArea) from the Global Map
of Irrigation Areas (GMIA) to consider the share of irrigated
area within a grid as an allocation parameter. GMIAv5.0 is
the only irrigated area dataset with global coverage, which
estimates the amount of area equipped for irrigation at a
5 arcmin resolution for the period around 2005 (Siebert et
al., 2013). GIMAv5.0 does not include information on the
functionality or quality of irrigation equipment and makes
no distinctions between different types of irrigation, which
may introduce errors and inconsistencies into the allocation.
We present a map of area equipped for irrigation at the grid
level (IrrArea
i
) in Fig. S3 in the Supplement (Sect. S9).
4.2.4
Protected area
We select the protected area (Protect) from the World
Database on Protected Areas (WDPA), released by the In-
ternational Union for Conservation of Nature (Deguignet et
al., 2014), as an allocation parameter to indicate the loca-
tions where crop production is least likely to take place. No-
tionally, crop production does not occur within protected ar-
eas (such as national parks, wilderness areas, and nature re-
serves), but in reality it does. During the initial allocation
process SPAM allows for crop allocation in protected areas
to allow for this reality, but if the model does not solve, one
option is to increase the area designated as cropland, suitable
land, or irrigated land. That expansion is not allowed into
protected areas. The data are originally in a polygon format.
We convert them to 5 arcmin grids (Protect
i
) and map them
in Fig. S4 in the Supplement (Sect. S10).
4.2.5
Accessibility
We adopt the population count from the Gridded Popula-
tion of the World (GPWv4.0) as a proxy to consider the
influence of market accessibility (Access) on farmers’ crop
choices (Eq. 10). GPWv4.0 provides a gridded representa-
tion of human populations across the globe at a 30 arcsec
resolution (CIESIN, 2016). For SPAM 2010, we aggregate
the population count grid to a 5 arcmin resolution and recal-
culate the population density. Then we derive rural popula-
tion density (AggRurPop
i
) based on the assumption that if
there is cropland within the 5 arcmin grids, then the popu-
lation residing within the grids should be rural people. We
do not aim to distinguish rural area from urban area. In-
stead, the variable AggRurPop
i
is introduced to estimate the
market accessibility and to account for subsistence produc-
tion. Therefore, it does not mean the accessibility of getting
food. As crop-specific revenue is divided by the total revenue
within a pixel (Eqs. 11 and 12), the prior is not affected by
market accessibility if it is not crop-specific. In other words,
crop-specific market accessibility is preferable for the cur-
rent SPAM model. Such accessibility does not exist now. We
create a measure of market accessibility (Access
i
) from the
grid-level estimates of rural population by considering the
relationship between AggRurPop
i
and maximum and mini-
mum rural population densities within a country. Population
in grids with no cropland is not used in further calculation.
We present the detailed procedure for measuring Access
i
in
the Supplement (Sect. S11), which further contains a map of
AggRurPop
i
(Fig. S5) and a table of minimum and maximum
rural population densities in select countries (Table S9).
4.2.6
Crop revenue
We measure the crop potential revenue (Rev) – determined
by market accessibility (Access), crop prices (Price), and
crop potential yield (PotYield) – as an allocation parameter,
which fully considers the influence of farmers’ crop choices.
We adopt the crop-specific prices (Price
j
) from the FAO’s
gross production value. Prices for crop aggregates (e.g., trop-
ical fruit) are calculated as a weighted average from FAO
world totals. It is important to note that these are not spa-
tially specific prices, and they likely misrepresent the local
economic realities and associated cropping choices faced by
farmers. We list the crop prices in Table S10 in the Sup-
plement (Sect. S12). We estimate the crop-specific potential
yield (PotYield
ij l
) as a composite measure of potential har-
vested yield (PotHarvYield
ij l
) based on GAEZ. We present
the detailed procedure for estimating PotYield
ij l
in the Sup-
plement (Sect. S12), which further contains a table listing the
dry-matter yield conversion factors (Table S11). Finally, we
calculate the grid-level potential unit revenue of planting a
certain crop according to Eq. (1).
4.3
Data harmonization and adjustment
4.3.1
Adjusting input data
We list the main input variables for SPAM2010 in Ta-
ble 2. As we collect data from various sources, it might in-
evitably cause information inconsistencies. Therefore, we set
rules to harmonize all these data. At the beginning, we ad-
just all the area-related parameters (e.g., cropland area, ir-
rigated area, and suitable area) to satisfy the constraints at
the administrative-unit level before calculating the priors of
physical area. When the model runs, it might be unable to
find the optimal allocation solution for a particular coun-
try, administrative unit, or crop. Under these circumstances,
we set several options to “force” a solution, including ad-
justing the entropy conditions and adjusting the data harmo-
nization rules. We elaborate on the details for adjusting areas
(Sect. S13), entropy conditions (Sect. S14), and harmoniza-
tion rules (Sect. S15), respectively, in the Supplement.
4.3.2
Adjusting allocation results
The model produces the allocated harvested area (AllocH
ij l
),
the allocated yield (AllocY
ij l
), and the production quantity
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 11
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3555
Table 2.
The main input variables used in SPAM2010.
Variables
Definition
Sources
k
Administrative-unit levels (
k
=
0
,
1
,
2)
GAUL, GADM
j
Crop type (Total
=
42)
FAOSTAT
l
Farming system (
l
=
I, H, L, S)
FAO reports etc.
CropHPY
Statistics on harvested area (H), production (P), and yield (Y)
FAOSTAT etc.
AvgCropHPY
j k
CropHPY averaged to 2009–2011
FAOSTAT etc.
AdjCropHPY
j k
AvgCropHPY
j k
averaged to 2009–2011 and scaled to FAO statistics
Developed by authors
Percent
j lk
Shares of farming systems
l
by crop
j
and administrative unit
k
.
Developed by authors
AdjCropH
j lk
Adjusted harvested area (H) by
j
,
l
, and
k
Developed by authors
AdjCropY
j lk
Adjusted yield (Y) by
j
,
l
, and
k
Developed by authors
AdjCropA
j lk
Physical area (A) by
j
,
l
, and
k
; AdjCropH
j lk
divided by CropIntensity
j lk
Developed by authors
CropIntensity
j lk
Harvesting frequency per year per unit cropland by
j
,
l
, and
k
Expert judgments etc.
i
5 arcmin grid cell
Developed by authors
AggMedCropLand
i
Median cropland in each grid
i
CAAS–IFPRI cropland
AggMaxCropLand
i
Maximum cropland in each grid
i
CAAS–IFPRI cropland
AggProbCropLand
i
Probability that estimated cropland amount is correct in each grid
i
CAAS–IFPRI cropland
AdjCropLand
i
Total cropland in each grid
i
, after adjustments
Developed by authors
SuitArea
ij l
Total suitable area in each grid
i
by crop
j
and farming system
l
GAEZv3.0
AdjSuitArea
ij l
Total suitable area in each grid
i
by crop
j
and farming system
l
, after adjustments
Developed by authors
IrrArea
i
Area equipped for irrigation in each grid
i
GMIAv5.0
AdjIrrArea
i
Total irrigated area in each grid
i
by crop
j
and farming system
l
, after adjustments
Developed by authors
Protect
i
Indicator of protected area in each grid
i
WDPA
Price
j
Prices for
j
calculated as a weighted average from world totals
FAO
AggRurPop
i
Population density in each grid
i
GPWv4.0
Access
i
Market accessibility in each grid
i
Developed by authors
PotHarvYield
ij l
Potential harvested yield from GAEZ in grid
i
by crop
j
and farming system
l
GAEZv3.0
PotYield
ij l
Potential yield calculated from PotHarvYield
ij l
in grid
i
by crop
j
and farming system
l
Developed by authors
Rev
ij l
Potential revenue in each grid
i
by crop
j
and farming system
l
Developed by authors
CropA
ij l
Prior allocation for physical area in each grid
i
by crop
j
and farming system
l
Developed by authors
π
ij l
Informed prior of physical area by
i
,
j
, and
l
, calculated from CropA
ij l
Developed by authors
s
ij l
Allocated shares of physical area in each grid
i
by crop
j
and farming system
l
Developed by authors
AllocA
ij l
Allocated physical area in each grid
i
by crop
j
and farming system
l
Developed by authors
Source:
developed by authors.
(AllocP
ij l
) for SPAM2010. As a final step, we need to ad-
just the allocation results in order to keep the grid-level re-
sults consistent with the statistics. In each step of estimation,
we scale the results to the national 2009–2011 FAO average
(AvgCropHPY
j k
0
) by crop
j
and country
k
0
to even out po-
tential inaccuracies introduced by the allocation adjustments.
This means all the allocated results in this subsection could
be adjusted (if necessary) before being applied in the next
phase.
We first scale the allocated harvested area (AllocH
ij l
) to
the national FAO average to even out potential inaccuracies
introduced by the allocation adjustments:
AdjAllocH
ij l
=
AllocH
ij l
P
i
∈
k
0
P
l
AllocH
ij l
×
AvgFAOCropH
j k
0
.
(19)
Total harvested area of each crop in the grid was calculated
by summing estimates across the four farming systems:
djAllocH
ij
=
X
l
AdjAllocH
ij l
∀
l.
(20)
For yield, we begin with the potential harvested yields
(PotHarvYield
ij l
) developed earlier (see Sect. S12 in the
Supplement). Missing values were filled in sequentially us-
ing the following values in order of availability.
i. Potential yield from potential-suitability surfaces:
AdjPotYield
ij l
=
PotHarvYield
ij l
;
(21)
ii. Average potential yield in SRU:
AdjPotYield
ij l
=
P
i
∈
k
SRU
PotHarvYield
ij l
×
AdjSuitArea
ij l
P
i
∈
k
SRU
AdjSuitArea
ij l
;
(22)
iii. Subnational yield by crop
j
, input
l
, and ADM2 unit
k
2
:
AdjPotYield
ij l
=
AdjCropY
j lk
2
;
(23)
iv. Subnational yield by crop
j
, input
l
, and ADM1 unit
k
1
:
AdjPotYield
ij l
=
AdjCropY
j lk
1
;
(24)
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 12
3556
Q. Yu et al.: A cultivated planet in 2010 – Part 2
v. National yield by crop
j
, input
l
, and ADM0 unit
k
0
:
AdjPotYield
ij l
=
AdjCropY
j lk
0
.
(25)
Then we modify the allocated yield (AllocY
ij l
) according
to the minimum and maximum yields in the administrative
unit:
ModAllocY
ij l
=
MinYield
j lk
SRU
if AllocY
ij l
<
MinYield
j lk
SRU
ModAllocY
ij l
=
MaxYield
j lk
SRU
if AllocY
ij l
>
MaxYield
j lk
SRU
ModAllocY
ij l
=
AllocY
ij l
if MinYield
j lk
SRU
≤
AllocY
ij l
≤
MaxYield
j lk
SRU
.
(26)
For production quantity, we scale the AllocP
ij l
to the national
FAO average:
AdjAllocP
ij l
=
AllocP
ij l
P
i
∈
k
0
P
l
AllocP
ij l
×
AvgCropP
j k
0
.
(27)
Then we calculate the total production in the grid by sum-
ming overall production levels:
AdjAllocP
ij
=
X
l
AdjAllocP
ij l
,
∀
l.
(28)
Finally, we recalculate the allocated yield from the allocated
harvested area and allocated production to effectively scale
yields to the national FAO average:
AdjAllocY
ij l
=
AdjAllocP
ij l
AdjAllocH
ij l
.
(29)
To simplify, the grid-cell yield is calculated from the reported
yield at the statistical reporting unit, the allocated area from
model results, and the potential yield (at grid-cell level) from
GAEZ (Global Agroecological Zone). These are illustrated
in Eqs. (15) and (16): the spatial variation in yield within
a statistical reporting unit follows the same spatial variation
in the potential yield of that crop. In other words, the more
suitable (higher potential yield) cells would have a relatively
higher yield while the average yield of all the grid cells would
be equal to the statistically reported yield of the administra-
tive unit.
5
Results
In this section, we briefly showcase some of the main
SPAM2010 results, which mainly focus on the staple crops,
to illustrate how SPAM2010 has been produced.
5.1
Disaggregated crop statistics
Disaggregation of crop statistics is the first step for running
the SPAM model. Table 3 summarizes the disaggregated rice
harvested area and yield (area-weighted) for global rice pro-
duction by four farming systems in SPAM2010. At the global
level, the world has harvested about 160
×
10
6
ha of rice
around 2010. The majority of rice production area is irri-
gated, i.e., about 98
×
10
6
ha, which accounts for 61.2 % of
the total rice harvested area. This share is followed by the
high-input rainfed farming system (17.3 %, approximately
27 million ha), subsistence farming system (16.0 %, approxi-
mately 26 million ha), and low-input rainfed farming system
(5.5 %, approximately 9 million ha). The global-average rice
yield is 4374 kg
/
ha, which stands at the average yield be-
tween the irrigated farming system (5528 kg
/
ha) and high-
input rainfed farming system (3663 kg
/
ha) and is much
higher than the average yield of low-input rainfed farming
system (1810 kg
/
ha) and the average yield of subsistence
farming system (1604 kg
/
ha). At the regional level, Asia
(South Asia, South East Asia, East Asia together) is the
largest rice-producing region, which has harvested approx-
imately 142 million ha of rice around 2010. The majority of
Asian rice production area is also irrigated, and the share, i.e.,
63.7 %, is close to the global share of irrigated rice farming
system. South Asia has more rice area harvested (approxi-
mately 60 million ha) than South East Asia (approximately
49 million ha) and East Asia (approximately 33 million ha).
However, the average rice yield in South Asia (3553 kg
/
ha)
is lower than South East Asia (4125 kg
/
ha) and East Asia
(6566 kg
/
ha). Consequently, the total rice production in these
regions is very close to each other. Rice production in North
America is completely irrigated, and the average yield is rel-
atively high in this region. Subsistence rice production is
mainly in sub-Saharan Africa (SSA) and South Asia, and
the rice yield under subsistence conditions is also the low-
est among the four farming systems.
5.2
Allocated harvested area and yield
After applying the optimization model in GAMS, the disag-
gregated crop statistics are spatially allocated to produce the
SPAM maps. Figures 3 and 4 present the maps of harvested
area and yield (after adjustment) for maize, respectively. For
all farming systems, as shown in Fig. 3e, maize area is highly
concentrated in northern China and North America. How-
ever, maize production in North America is mainly rainfed
with high input, while in China, the rainfed farming system
is mainly located in the northeastern part (Fig. 3e), and the
irrigated farming system is mainly found in the north-central
part (Fig. 3a). The rainfed low-input farming system (Fig. 3c)
and subsistence farming system (Fig. 3d) for maize produc-
tion are mainly located in South America and SSA, while
the rainfed high-input maize farming system is also widely
distributed outside China and northern America, including
Central America, Europe, and other regions (Fig. 3b). As
shown in Fig. 4e, the average maize yield is very high in
North America and Europe and is relatively high in South
America and Asia.
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 13
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3557
Table 3.
Regional values for area and yield of rice from SPAM2010. Unit: area (1000 ha); yield (kg
/
ha).
Region
Irrigated
Rainfed–high
Rainfed–low
Subsistence
Total
Area
Yield
Area
Yield
Area
Yield
Area
Yield
Area
Yield
North America
1259
7779
–
–
–
–
–
–
1259
7779
Central America
624
4408
147
2582
150
2228
49
2492
969
3698
South America
2126
7510
536
3320
1261
2536
995
2691
4917
4803
Europe
540
6352
146
3315
122
7531
6
5340
814
5977
Middle East & North Africa
1034
6848
–
–
114
4340
–
–
1148
6599
SSA
2172
3844
415
3542
2513
1653
4413
1412
9514
2124
South Asia
32 156
4572
11 122
3536
3755
1405
12 746
1628
59 779
3553
South East Asia
25 620
5170
14 954
3774
850
1745
7500
1522
48 924
4125
East Asia
32 609
6610
561
4037
2
626
–
–
33 173
6566
Russia
195
5173
–
–
–
–
–
–
195
5173
Oceania
34
9620
–
–
5
3242
0
1960
39
8665
World
98 369
5528
27 881
3663
8773
1810
25 709
1 604
160 732
4374
Source
: developed from our own calculations.
Figure 3.
Harvested-area maps for maize in irrigated
(a)
, rainfed–high
(b)
, rainfed–low
(c)
, subsistence
(d)
, and all
(e)
farming systems.
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 14
3558
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Figure 4.
Yield maps for maize in irrigated
(a)
, rainfed–high
(b)
, rainfed–low
(c)
, subsistence
(d)
, and all
(e)
farming systems.
6
Value of production
Finally, we use the average 2009–2010 base year price (in-
ternational dollar, I$) to compute value of production in each
grid and for each crop and farming system. Table 4 shows
value of production for all crops, food and nonfood crops in
all regions, and the percentage of each category value in re-
lation to the total value. Asia (South Asia, South East Asia,
East Asia together) accounts for nearly half (49.2 %) of the
total value of crop production in 2010, while the Middle East
and North Africa, Central America, Russia, and Oceania ac-
count for less than 5 % each. Globally, food crops account
for 86.2 % of the total crop production value, with minor re-
gional differences (the classification of crops into food and
nonfood is detailed in Table S3).
7
Data availability
The SPAM2010 provides four essential output indicators, in-
cluding the following.
a. PHYSICAL AREA: it is measured in hectares and rep-
resents the actual area where a crop is grown, not count-
ing how often production was harvested from it. Phys-
ical area is calculated for each production system and
crop, and the sum of all physical areas of the four pro-
duction systems constitutes the total physical area for
that crop. The sum of the physical areas of all crops in
a pixel may not be larger than the pixel size.
b. HARVESTED AREA: also measured in hectares, har-
vested area is at least as large as physical area but some-
times more since it also accounts for multiple harvests
of a crop on the same plot. Like for physical area, the
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 15
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3559
Table 4.
Value of production for all crops as well as food and nonfood crops in various regions.
Region
All crops
Food crops
Nonfood crops
(million I$)
(million I$)
(percent)
(million I$)
(percent)
North America
139 173
125 655
0.90
13 518
0.10
Central America
33 174
26 340
0.79
6834
0.21
South America
135 222
99 950
0.74
35 272
0.26
Europe
202 180
165 761
0.82
36 419
0.18
Middle East & North Africa
60 643
51 358
0.85
9285
0.15
SSA
110 406
96 474
0.87
13 931
0.13
South Asia
201 196
171 814
0.85
29 382
0.15
South East Asia
129 189
105 878
0.82
23 311
0.18
East Asia
367 338
344 799
0.94
22 539
0.06
Russia
26 489
22 782
0.86
3707
0.14
Oceania
13 426
11 337
0.84
2089
0.16
World
1 418 435
1 222 147
0.86
196 288
0.14
Source
: developed from our own calculations.
harvested area is calculated for each production system,
and the sum of all harvested areas of all production sys-
tems in a pixel amounts to the total harvested area of the
pixel. The sum of all the harvested areas of the crops in
a pixel can be larger than the pixel size.
c. PRODUCTION: for each production system and crop,
production is calculated by multiplying area harvested
by its corresponding yield. It is measured in metric tons.
The total production of a crop includes the production
of all production systems of that crop.
d. YIELD: it is a measure of productivity, the amount of
production per harvested area, and is measured in kilo-
grams per hectare. The total yield of a crop, when con-
sidering all production systems, is not the sum of the
individual yields but the weighted average of the four
yields.
The SPAM2010 can be downloaded from the Har-
vard Dataverse (https://doi.org/10.7910/DVN/PRFF8V; IF-
PRI, 2019), which includes all results of maps, tables,
and figures. Registered users can find more informa-
tion for the SPAM model and the previous versions of
SPAM datasets via the dedicated MapSPAM website (http:
//mapspam.info/, last access: 11 December 2020). The for-
mal SPAM products in 2000 and 2005 are also avail-
able on the MapSPAM website. Their Dataverse addresses
are https://doi.org/10.7910/DVN/A50I2T (SPAM, 2000) and
https://doi.org/10.7910/DVN/DHXBJX (SPAM, 2005). All
these three datasets are in the same place grouped un-
der IFPRI Harvest Choice Dataverse (https://dataverse.
harvard.edu/dataverse/harvestchoice, last access: 11 Decem-
ber 2020).
8
Discussion
8.1
Model uncertainty and validation
The first SPAM product was the regional-level agricultural-
production maps produced for Brazil circa 1994 (You et al.,
2006). Since then the model and products of SPAM have
been continuously improved and updated. Besides the evo-
lution of the method (see Sect. 3), the evaluation of SPAM
model performance is also improving. In one of our early
works, the uncertainty in the model, i.e., the variance ex-
plained by the cross-entropy approach, is evaluated by com-
paring it with the performance of simplified proportional ap-
proaches, which have been used by Monfreda et al. (2008)
for producing the M3 dataset. It was proven that the cross-
entropy approach was more successful in estimating crop
areas than the proportional approaches, no matter the pro-
portion to the total land area, to the cropland area, or to
the amount of (biophysically) suitable land for the produc-
tion of each crop (You et al., 2006). Moreover, many re-
searchers believed that the inclusion of economic factors,
i.e., market, would increase the performance of the crop dis-
aggregation model (see the discussion in You et al., 2014),
though it does not automatically guarantee that model’s out-
puts. This partly explained the considerable discrepancies be-
tween SPAM2000 and M3 (Anderson et al., 2015) and partly
confirmed that using more sophisticated approaches for pro-
duction allocation would reduce uncertainty (Donaldson and
Storeygard, 2016). In one of our recent works, the sensitivity
of the variant of the standard SPAM model output to a few
methodological-data choices had been evaluated. These in-
clude the spatial allocation method, the crop coverage, the
treatment of a “rest-of-crops” aggregate, the incorporation
of a “crop potential suitability” data layer, the inclusion of
rudimentary economic elements, and the administrative-unit
details of the primary source statistics. It showed that the
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 16
3560
Q. Yu et al.: A cultivated planet in 2010 – Part 2
standard SPAM estimates are unsensitive to the inclusion of
crude economic elements, moderately sensitive to the set of
crops or crop aggregates being modeled, and mostly depen-
dent on the degree of disaggregation of the underlying na-
tional and subnational production statistics (Joglekar et al.,
2019). This implies that the improvements to the method-
ological aspect of SPAM have limited effect on reducing un-
certainty. By contrast, the quality and accuracy of the un-
derlying statistics used to prime the model are particularly
pertinent (Joglekar et al., 2019).
SPAM products are estimates with various uncertainty. In-
accuracy surely exists and varies from region to region and
even from crop to crop. Although there were efforts paid to
the evaluation of model conceptualization and performance,
those previous validations should not be taken for granted
for the latest updates. Therefore, we carried out extensive
validation work to assess the accuracy of the output maps
of SPAM2010. Firstly, we relied on a system through which
we are able to send the crop maps to collaborators and users
alike for comments or assessment. For example, the CGIAR
is a global partnership which unites 15 centers engaged in
agricultural research. Each center has its own mandate crops,
e.g., IRRI (International Rice Research Institute) for rice
and CIMMYT (International Maize and Wheat Improvement
Center) for maize and wheat. We took advantage of their vast
network of field offices and local expertise to help us to val-
idate the SPAM results. Many researchers from these insti-
tutes have been involved in the production of SPAM2010,
which increases the reliability of the results. The Chinese
Academy of Agricultural Sciences (CAAS) undertook the
regional-level validation for SPAM2010 following the ap-
proaches they have applied for the evaluations of previous
SPAM products (Liu et al., 2013; Li et al., 2016; Chen et al.,
2016). Moreover, field-level validating information has ei-
ther been collected by crowdsourcing tools such as Geo-Wiki
(Fritz et al., 2012) and eFarm (Yu et al., 2017a) or through
field trips and workshops on-site or online, where local ex-
perts were asked to confirm or validate the crop production
maps by providing handwritten comments or posting com-
ments online at the our MapSPAM website. Most of these
reports were collected crop by crop and country by country.
An example of the detailed validation process is provided
in the Supplement (see Sect. S16). The complete validation
process could take a great deal of effort and time, but these
users’ feedback is quite important and valuable. We took this
feedback and rerun SPAM model and further released the
updated versions of SPAM. The previous SPAM products
have been updated substantially with the help of those com-
ments. For example, SPAM2000 and SPAM2005 are at ver-
sion 3.07 and version 3.20, respectively. The current product,
i.e., SPAM2010v1.10, was already released after extensive
validations; it is still open and ready to receive more com-
ments. Such an iterative process would enable a continued
update to improve the product quality.
Secondly, we qualitatively evaluated the uncertainty in the
input data. Like any models, the results depend on the input
data and the modeling process. For SPAM, the most impor-
tant input data are the subnational crop data, which have a
large impact on the final product accuracy as mentioned be-
fore. We built our SPAM uncertainty rating mainly on the
availability and confidence of our subnational data. In addi-
tion, we added the parameters and constraints we have to ad-
just to solve the SPAM model. For example, we sometimes
have to abandon some crop-potential-suitability constraints
in order to solve a country. For some countries, we may have
to allow cropland per pixel to increase by 5 % or even 10 %
of the original input to make the model run. In addition, we
collected feedback and comments from users, local experts,
and collaborators as discussed above. They are sporadic but
very useful. We combine all the information together to give
a subjective rating on how confidence we, the SPAM team,
think of our final crop maps (both area and yield) based on
the judgment of the reliability of input data. Figure 5 shows
the country-level uncertainty rating with 5 categories (1 rep-
resents the lowest uncertainty, 5 the highest). The complete
rating list is presented in Sect. S17 in the Supplement. Not
surprisingly, the uncertainty in Africa and South East Asia
is higher than that in countries in Europe and America. Al-
though such a validation process is not vigorous, the result
is convincing, and such a rating is highly demanded and ex-
plicitly requested by users.
Thirdly, we quantitatively evaluated the results by cross-
comparing the results with statistics at another administrative
level that has not been used in running the model. We ran
SPAM with complete statistics (ADM0, ADM1, and ADM2)
and then ran them with only ADM0 and ADM1 statistics to
see how the aggregated results to ADM2 compare to the orig-
inal statistics at ADM2, or at least to the aggregated origi-
nal results at ADM2. The runs were all done at ADM1 and
then combined to give results for the whole country. We then
calculated the coefficient of determination (
R
2
) between the
values allocated from model and obtained from statistics to
assess the model performance. In general, a higher
R
2
indi-
cates a better performance. This approach has already been
used for evaluating the performance of SPAM2000 (You et
al., 2014). The upper part of Fig. 6 shows the results of
such an approach applied to Brazil in SPAM2000 for its
main food crops, while the bottom part of Fig. 6 shows the
results of the same approach applied to the same country
for the same crops in SPAM2010. The figure clearly indi-
cates that the model performed better in allocating rice than
other crops. Moreover, the performance improved greatly
from SPAM2000 to SPAM2010, especially for soybean and
potato. We further selected a few smaller countries in Asia
and Africa to undertake the same assessment, which are be-
lieved to have a relatively higher uncertainty in terms of in-
put data (Fig. 5). Bangladesh, Benin, Senegal, and Tanza-
nia were selected as they have good statistical data cover-
age in SPAM2010. Figure 7 shows that the
R
2
for selected
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 17
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3561
Figure 5.
Subjective uncertainty rating for SPAM2010 input data by individual countries.
crops (i.e., maize, rice, and cotton) ranged between 0.66 and
0.94, suggesting that the overall performance of SPAM2010
is good in these selected countries for those selected crops.
Finally we did regional-level quantitative validations when
third-party independent crop maps were available, given that
it is impossible for us to collect the true spatial distribution
of crops (both area and yield) for the time of 2010 on a
global scale. Among the limited third-party independent spa-
tial crop distribution data, the Cropland Data Layer (CDL;
https://nassgeodata.gmu.edu/CropScape/, last access: 11 De-
cember 2020) is a crop-specific land cover dataset created
for the continental United States using moderate-resolution
satellite imagery and extensive agricultural ground truth,
which has been applied to validate our SPAM2010 product
at the regional scale by correlating the grid-level crop area.
We focus on the three most popular staple crops in the United
States, i.e., maize, wheat, and soybean, and obtain the crop
area maps of 2009, 2000, and 2011 from CDL. We calcu-
late the 2009–2011 average crop areas at a 5 arcmin reso-
lution for CDL according to the scheme of SPAM2010 and
further calculate the coefficient of determination (
R
2
) and
the root mean square error (RMSE) between the grid-level
values derived from the two datasets (Fig. 8). The values
of
R
2
are between 0.71 and 0.91, and the values of RMSE
are between 231 and 307 ha, indicating a relatively high reli-
ability. In particular, the higher
R
2
and lower RMSE sug-
gest our maize and soybean maps are more reliable than
the wheat map. There are potentially many factors affecting
the different results if we treat CDL as the truth, for exam-
ple, the different accuracy or availability of input data, suit-
ability layers, and parameters for the area shares and yield
ratios. Another possible reason is that we did not distin-
guish spring wheat and winter wheat in SPAM, which partly
explains why the agreement for wheat is lower than that
for maize and soybean. Moreover, the National Land Cover
Dataset (NLCD) of China mapped paddy field distribution
as a special cropland cover at a 1
×
1 km grid level (http:
//www.resdc.cn/data.aspx?DATAID=99, last access: 11 De-
cember 2020). By assuming paddy fields will be mostly used
for growing rice, we evaluate the rice area map in China
by correlating SPAM2010_rice and NLCD2010_paddy ac-
cording to the same scheme described above. The value of
R
2
is 0.49, and the value of RMSE is 1024 ha (Fig. 9). Al-
though this result does not seem as good as the results from
the United States by using CDL, it is fairly acceptable be-
cause NLCD measures land cover rather than land use and
is in a relatively coarse spatial resolution. Moreover, the
R
2
is substantially increased comparing to its predecessors. For
example, the
R
2
is assessed as 0.42 for SPAM2005 by using
the same approach according to Liu et al. (2013). In addition,
there are regional-level crop distribution maps produced by
independent efforts on interpreting remotely sensed images.
For example, Zhang et al. (2017) provided annual paddy area
time series from 2000 to 2010 based on satellite remote sens-
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 18
3562
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Figure 6.
Comparison between the allocated crop area and statistics crop area at the ADM2 level in Brazil (log–log scale plot; unit: ha). The
upper part is for SPAM2000 and the bottom part is for SPAM2010.
ing for China and India. We compared these remote-sensing-
derived paddy maps with the rice area estimated by SPAM
for the year 2010. The
R
2
values are 0.36 and 0.34 for China
and India, respectively (Fig. 10). We could expand this quan-
titative evaluation when more third-party independent crop
maps are available. However, it should be noted that errors
might exist in the third-party independent crop maps as well;
hence this quantitative evaluation approach also might result
in uncertainty. Our results show that the uncertainty grad-
ually increases when applying CDL, NLCD, and Zhang et
al. (2017).
8.2
Data comparison
There are a few reports which compare SPAM with M3,
MIRCA, and GAEZ, especially their output maps circa 2000
(Anderson et al., 2015; Donaldson and Storeygard, 2016).
Although it is difficult to make statements about which one is
better, there are several features that distinguish SPAM prod-
ucts from the M3, MIRCA, and GAEZ data. First, the es-
timates from SPAM can be customized using user-provided
data for one or more of the input variables and return re-
sults to the provider in a short turnaround period. Second,
although SPAM runs mainly at a 5 arcmin resolution, it can
be run at higher resolutions provided that at least some of
the rasterized inputs also have higher-resolution data to sup-
port such an exercise. Third, considerable effort is made to
compile subnational crop statistics at administrative level 2
(e.g., district or county) for all possible countries. Fourth, if
there is knowledge of crop existence in any area for any crop,
this can be incorporated into the model to make a more accu-
rate crop allocation. Moreover, SPAM does not have a large
coverage of crops (compared to M3) and does not include
detailed biophysical parameters (compared to MIRCA and
GAEZ); instead it focuses more on agricultural production
by providing data on crop harvested area and yield disaggre-
gated by farming systems. Finally, SPAM results are readily
available on the internet in several formats (also tabular) for
all interested users. We are currently building a SPAM model
in the cloud, where we let any user supply his or her own in-
put data and run SPAM on his or her own under the GitHub
platform. This SPAM in the cloud will be published and com-
municated to the SPAM user community once it is ready.
Anderson et al. (2015) conclude that substantial discrepan-
cies exist across these four global spatially explicit crop pro-
duction datasets circa 2000, and the disagreement between
models serves as a reminder of the ongoing challenges to
the creation of spatially explicit estimates of harvested area
and yield based on crop statistics. However, it is more chal-
lenging to assess the disaggregated farming system results
such as irrigated rice vs. rainfed rice and subsistence maize
vs. high-input rainfed rice, which have not been systemati-
cally explored in Anderson et al. (2015). We collected ad-
ditional global datasets which are relevant to agricultural-
production mapping, e.g., the average irrigated and rainfed
yields (ca. 2000) from Siebert and Döll (2010) and the har-
vested area and yield for four crops (ca. 2005) from http:
//www.earthstat.org/ (last access: 11 December 2020). We
compared these datasets with our SPAM products for the
corresponding period. We found that the results are differed
from crop to crop and from farming system to farming sys-
tem. In general, the yield estimates on maize and wheat are
better than the other crops, and the irrigated yields are better
than the rainfed yields (Figs. 11 and 12).
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 19
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3563
Figure 7.
Comparison between the allocated crop area and statistics crop area at the ADM2 level in Bangladesh, Benin, Senegal, and
Tanzania for maize, rice, and cotton (log–log scale plot; unit: ha).
However, as M3, MIRCA, and GAEZ do not provide sub-
sequent global spatially explicit crop production ca. 2000,
it is impossible to compare the current SPAM2010 with
other data products. In order to illustrate the continuity of
SPAM products, we present a grid-by-grid comparison be-
tween SPAM2010 and SPAM2005. Figure 13 shows that
rice production in 2010 increased notably in eastern Eu-
rope, Africa, northeastern China, northwestern India, south-
ern Australia, etc., while it decreased notably in central Asia
and South America. Maize production displays an overall
increase across the globe between 2005 and 2010, except
for some places in central Asia, which have shown a de-
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 20
3564
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Figure 8.
Grid-by-grid comparison of crop area for maize
(a)
, wheat
(b)
, and soybean
(c)
between SPAM2010 and CDL2010 in the conti-
nental US.
Figure 9.
Grid-by-grid comparison between SPAM2010 rice area
and NLCD2010 paddy field area in China.
creasing trend. It is also noticeable that maize production
in the US and Europe has kept relatively stable. This re-
sult is accordant with the “maize boom” which had taken
place around the globe (Herrmann, 2013), especially in the
developing countries (Cairns et al., 2013; Ornetsmüller et al.,
2019). It should be noted that the current type of comparison
may not be a perfect comparison (because differences exist
in methodologies and data input applied in SPAM2005 and
SPAM2010) and that the current comparison only shows the
rate of change; thus a higher value does not necessarily indi-
cate a huge change in absolute crop production.
In addition, we compared the changes in crop area between
SPAM products and the abovementioned regional-level inde-
pendent crop maps once they were available in time series.
We calculated the area changes in maize, wheat, and soy-
bean by overlaying CDL2005 and CDL2010 and undertook
the same procedure for SPAM. We then plot these changes
(i.e.,
1
CDL and
1
SPAM) in Fig. 14. Likewise, we compared
the changes in SPAM rice area and the changes in paddy
rice area obtained from Zhang et al. (2017; Fig. 15). Fig-
ures 14 and 15 both show that the coefficient of determina-
tion is extremely low between changes yielded from different
data products, which further reminds us that it is inappropri-
ate to directly compare SPAM products over time, although
we are confident of the spatial accuracy of SPAM products
at each time stage. This is mainly because SPAM requires a
large number of input data, yet the sources of these multiple
data inputs cannot be guaranteed to be the same across differ-
ent time stages. Therefore, such changes reflected by SPAM
products over time not only mix real changes on the ground
but also largely depend on the input data. For example, the
cropland layers (one of the most important data inputs) are
accessed from different sources to make sure the cropland
data and the statistical data are adopted for the same year. We
did not evaluate the continuity of these input data, which is
almost impossible and is beyond the purpose of SPAM. Con-
sequently, it is suggested to use the SPAM products with, at
least, acknowledgement to the corresponding cropland layer,
e.g., Lu et al. (2020) for SPAM2010. Moreover, we do not
recommend users to cross-compare the SPAM products over
time because the differences may have more input data errors
or inaccuracies than detecting the real change on the ground.
8.3
Limitations
As stated previously, the SPAM estimates are dependent on
the extent and veracity of the primary input data like most
models (Joglekar et al., 2019). SPAM2010 requires data on
42 crops in over 200 countries for the production season. Ide-
ally, these data should be collected at an ADM2 level; how-
ever this is not always possible. It is particularly difficult to
a few countries such as Somali and Nigeria, where reliable
data are not available, or different input data just conflict with
each other. For example, only one crop area (i.e. millet) for
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 21
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3565
Figure 10.
Grid-by-grid comparison between SPAM2010 and Zhang et al. (2017) rice area in China and India.
a district is already larger than the total cropland area, yet
we know there are still five more crops growing in this dis-
trict. In these cases, we have to adjust the conflicting data
using expert judgment to make the model solvable. Since
most cropping statistics are not delineated by farming sys-
tem, estimates of the shares of production under each of the
four systems in question are required. To convert harvested-
area statistics to the physical-area statistics used in the model,
additional data on cropping intensities by crop and farming
systems must be collected. We have made every effort to col-
lect official or published data, and we reply on expert judg-
ments as the last resort when we simply could not find other
sources. For example, no country publishes official statistics
on crop yield ratio (yield conversion factor) between irrigated
vs. rainfed crops. We surveyed published papers, personal
communication with the FAO’s Agriculture to 2030 team,
and gray literature to collect such data. While indeed a se-
ries of expert judgments are used, the scope (e.g., crops and
regions) is quite limited in the overall input data. Once the
data on disaggregate cropping practices are compiled, several
variables at a gridded scale are needed to disaggregate these
cropping statistics into the desired spatial units. These data
include estimates of cropland, irrigated land, suitable area
and yield, population density, and protected areas. The vari-
ety and sheer volume required to run the SPAM (and related)
models raise questions of reliability and comprehensiveness
of estimates across different cropping statistics, geographic
areas, and countries.
In terms of reliability, different sources of information may
lead to inconsistent and even incompatible information. For
example, the data on the estimated cropland extent within
a grid are compiled from several sources, which in turn de-
ploy different methods to generate their estimates. The ex-
tent of cropland within a grid is crucial information for the
allocation model, but the confidence regarding its actual lo-
cation varies regionally (see Lu et al., 2020). Crop statistics
on area harvested and yield may not have been consistently
collected and processed across different countries, so these
major data may be unreliable to begin with. Additionally,
two of the major conversion factors used, farming system
shares and cropping intensities, are often not available for
each crop and farming system within a country. Lacking raw
data on these statistics for a particular crop–country combi-
nation, these data were simply assigned from a similar crop
or country or created using expert judgment. Neither data on
cropping intensities nor farming system shares have been val-
idated for reliability. In terms of comprehensiveness, notably
less subnational coverage exists in developing countries, and
only global-average commodity price data were used to ac-
count for the economic influences on crop production.
The wide range of data sources, coverage, and regional
nuances of crop production have methodological implica-
tions. First, there are possible trade-offs between data con-
sistency and data reliability. For example, there are require-
ments of the model (i.e., cropping intensities and farming
system shares) that are not consistently available within a
country at the administrative level needed. Often, these num-
bers are taken from national-level values, even though they
may not reflect the reality at the administrative level. Sec-
ondly, multi- and intercropping are not handled in a sophisti-
cated manner within SPAM. These types of cropping patterns
are only accounted for using a single cropping intensity value
per crop, farming system, and (possibly) subnational unit.
Finally, the market is important for both subsistence farm-
ers and commercial ones (Losch et al., 2012). In SPAM, we
rely on population density for an indirect representation of
market proximity. We use the rural population to calculate a
prior for the
subsistence portion
of a crop (i.e., subsistence
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 22
3566
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Figure 11.
Grid-by-grid comparison between SPAM2000 and Siebert and Doll (2010) in average irrigated and rainfed yields (log–log scale
plot; unit: kg
/
ha).
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 23
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3567
Figure 12.
Grid-by-grid comparison between SPAM2005 and EARTHSTAT2005 in crop yields. (log–log scale plot; unit: kg
/
ha).
among the four farming systems). For subsistence farmers,
by definition they mostly consume what they produce, and so
we indeed assume their production is closely correlated with
their rural population size. For all other farmers, we assume
they produce for the market (local, regional, or even inter-
national markets). The assumption of using rural population
as a proxy for market access was to break down the total
cropland into individual crop areas in the prior revenue cal-
culation. We did not, explicitly or implicitly, assume a rural
population (even subsistence farmers) is entirely fed by local
agriculture. Instead, we applied rural population density as a
proxy for market access, which allows crop-specific tuning
depending on how crops are distributed across the four farm-
ing systems. As this crop-specific revenue is divided by the
total revenue within a pixel in Eqs. (9) and (10), the prior is
not affected by market accessibility if it is not crop-specific.
In other words, crop-specific market accessibility is prefer-
able for the current SPAM model. Yet such accessibility data
at the global scale do not exist now. We would consider in-
cluding a more direct representation of market proximity by
actual travel times to markets or road networks as the global
roads and railways databases are becoming available (Koks
et al., 2019). Several trade-offs were made to ensure that the
complex allocation method was tractable, and it is important
to recognize that these trade-offs likely affect the plausibility
of results.
Last but not least, we admit that it is inappropriate to com-
pare SPAM products directly across time stages (Figs. 13–
15), although we have paid every effort to guarantee the
spatial accuracy of SPAM products at each time stage. It is
largely because the system errors exist across various data
products. In a latest publication, Iizumi and Sakai (2020) re-
leased a time series product of global gridded crop yields.
Although they applied a different approach (i.e., spatial ad-
justment) which is conceptually different from the spatial
disaggregation approach applied in SPAM, it provides great
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 24
3568
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Figure 13.
Comparison between SPAM2010 and SPAM2005:
(a)
relative difference in rice production,
(b)
relative difference in maize
production.
Figure 14.
Comparison between SPAM crop area change and CDL crop area change (log–log scale plot; unit: ha).
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 25
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3569
Figure 15.
Comparison between SPAM rice area change and Zhang et al. (2017) paddy rice change (unit: ha).
implications to further integrate and standardize the SPAM
and the similar gridded earth system datasets for broader ap-
plications. There is an ongoing consortium called The Land
Use Change Knowledge Integration Network (LUCKiNet;
http://www.luckinet.org, last access: 11 December 2020).
The SPAM team is part of this consortium, which aims to
integrate tools and standardize approaches across various on-
going projects that develop gridded information on land use
dynamics for applications in food security, climate change,
biodiversity, and other related issue areas. Not only LUCK-
iNet aims to create crop maps comparable over time, but we
also want to have these maps consistent across land uses such
as cropland, grassland, and forest. The modeling techniques
would consider the spatiotemporal dynamics of different land
use forms in an integrative framework.
8.4
Concluding remarks
In this paper, we present SPAM2010, the latest global grid-
ded agricultural-production dataset in 2010. SPAM2010 uses
an updated cross-entropy approach to make plausible esti-
mates of crop distribution for 42 crops and four farming sys-
tems within disaggregated units, which shows great improve-
ment over its predecessors: SPAM2000 and SPAM2005. For
example, the expanded crop list enables the analysis for not
only staple food crops but also cash crops. A recent study has
analyzed the global beer supply by using SPAM2000 (Xie et
al., 2018). It will be very promising to analyze the global cof-
fee and tea supply by using the latest dataset as these crops
are newly included in SPAM2010 and are in increasing de-
mand with superior economic value but also highly sensitive
to climate change (Bunn et al., 2015).
SPAM2010 substantially extends the SPAM family and
fills the gap for the work of global agricultural-production
mapping by successfully creating a global gridscape at the
confluence between earth and farming systems. In partic-
ular, it helps to better understand land management prac-
tices characterized by concomitant data and knowledge gaps
(e.g., crop selection and element of crop harvest; Erb et al.,
2017). It not only allows analysts and policymakers to bet-
ter target agricultural and rural development policies and in-
vestments, increasing food security and growth with minimal
environmental impacts, but also enables scientists to better
address the global-change challenges within the anthropo-
sphere and beyond by providing the only possibility for up-
dating the global agricultural and environmental assessments
from 2000 (when M3, MIRCA, GAEZ, and SPAM2000
are available) to 2005 and 2010 (when SPAM2005 and
SPAM2010 are available as well). All the SPAM maps and
tabular data in multiple time stages are freely available on
the MapSPAM website (http://mapspam.info/ last access:
11 December 2020), which also acts as a platform for val-
idating and improving the performance of the SPAM maps
by collecting feedback from users.
Supplement.
The supplement related to this article is available
online at: https://doi.org/10.5194/essd-12-3545-2020-supplement.
Author contributions.
QY, LY, WW, and PY framed the work.
QY, LY, UWS, and YR developed the SPAM2010 dataset. QY, LY,
ML, WW, and PY developed the cropland layer, which has been
used as the main input data for SPAM2010. SF and WX helped
the validation of SPAM2010. UWS and AKBJ prepared a technical
document. QY prepared the manuscript with contributions from all
the coauthors and wrote the final paper.
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 26
3570
Q. Yu et al.: A cultivated planet in 2010 – Part 2
Competing interests.
The authors declare that they have no con-
flict of interest.
Acknowledgements.
The work forms part of the CAAS Agri-
cultural Science and Technology Innovation Program (ASTIP),
CGIAR Research Program on Water, Land and Ecosystems led by
International Water Management Institute (IWMI), the CGIAR Re-
search Program on Policies, Institutions, and Markets (PIM) led by
International Food Policy Research Institute (IFPRI), and the Inter-
national Science Technology Practice and Policy Center and GEMS
Agroinformatics Initiative at the University of Minnesota. All peo-
ple and institutes provided data or comments or feedback to SPAM
are acknowledged. This work contributes to the Global Land Pro-
gramme (http://www.glp.earth, last access: 11 December 2020).
Financial support.
This research has been supported by the Na-
tional Natural Science Foundation of China (grant nos. 41921001
and 41871358), the National Key Research and Development
Program of China (grant no. 2017YFE0104600) and the Cen-
tral Public-interest Scientific Institution Basal Research Fund (no.
Y2020YJ07).
Review statement.
This paper was edited by Yuyu Zhou and re-
viewed by Patrick Meyfroidt and Toshichika Iizumi.
References
Anderson, W., You, L., Wood, S., Wood-Sichra, U., and Wu, W.:
An analysis of methodological and spatial differences in global
cropping systems models and maps, Global Ecol. Biogeogr., 24,
180–191, https://doi.org/10.1111/geb.12243, 2015.
Bonan, G. B. and Doney, S. C.: Climate, ecosystems, and planetary
futures: The challenge to predict life in Earth system models, Sci-
ence, 359, eaam8328, https://doi.org/10.1126/science.aam8328,
2018.
Bunn, C., Läderach, P., Ovalle Rivera, O., and Kirschke, D.:
A bitter cup: climate change profile of global production of
Arabica and Robusta coffee, Climatic Change, 129, 89–101,
https://doi.org/10.1007/s10584-014-1306-x, 2015.
Cairns, J. E., Hellin, J., Sonder, K., Araus, J. L., MacRobert, J. F.,
Thierfelder, C., and Prasanna, B. M.: Adapting maize production
to climate change in sub-Saharan Africa, Food Sec., 5, 345–360,
https://doi.org/10.1007/s12571-013-0256-x, 2013.
Chen, H., Li, Z., Tang, P., Hu, Y., Tan, J., Liu, Z., You, L., and Yang,
P.: Rice area change in Northeast China and its correlation with
climate change, J. Appl. Ecol., 27, 2571–2579, 2016.
Chen, J., Chen, J., Liao, A., Cao, X., Chen, L., Chen, X., He,
C., Han, G., Peng, S., Lu, M., Zhang, W., Tong, X., and Mills,
J.: Global land cover mapping at 30 m resolution: A POK-
based operational approach, ISPRS J. Photogramm., 103, 7–27,
https://doi.org/10.1016/j.isprsjprs.2014.09.002, 2015.
CIESIN: Gridded Population of the World, Version 4 (GPWv4):
Population Count Adjusted to Match 2015 Revision of UN WPP
Country Totals, Center for International Earth Science Informa-
tion Network, Columbia University, USA, 24 pp., 2016.
Deguignet, M., Juffe-Bignoli, D., Harrison, J., MacSharry, B.,
Burgess, N. D., and Kingston, N.: United Nations List of Pro-
tected Areas, UNEP World Conservation Monitoring Centre,
Cambridge, UK, 30 pp., 2014.
Deutsch, C. A., Tewksbury, J. J., Tigchelaar, M., Battisti, D. S.,
Merrill, S. C., Huey, R. B., and Naylor, R. L.: Increase in crop
losses to insect pests in a warming climate, Science, 361, 916–
919, https://doi.org/10.1126/science.aat3466, 2018.
Donaldson, D. and Storeygard, A.: The View from Above: Appli-
cations of Satellite Data in Economics, J. Econ. Perspect., 30,
171–198, https://doi.org/10.1257/jep.30.4.171, 2016.
Ehrensperger, A., de Bremond, A., Providoli, I., and Messerli, P.:
Land system science and the 2030 agenda: exploring knowl-
edge that supports sustainability transformation, Curr. Opin. Env.
Sust., 38, 68–76, https://doi.org/10.1016/j.cosust.2019.04.006,
2019.
Ellis, E. C., Kaplan, J. O., Fuller, D. Q., Vavrus, S.,
Klein Goldewijk, K., and Verburg, P. H.: Used planet: A
global history, P. Natl. Acad. Sci. USA, 110, 7978–7985,
https://doi.org/10.1073/pnas.1217241110, 2013.
Erb, K.-H., Luyssaert, S., Meyfroidt, P., Pongratz, J., Don,
A., Kloster, S., Kuemmerle, T., Fetzel, T., Fuchs, R.,
Herold, M., Wiltshire, A., and Dolman, A. J.: Land man-
agement: data availability and process understanding for
global change studies, Glob. Change Biol., 23, 512–533,
https://doi.org/10.1111/gcb.13443, 2017.
Erb, K.-H., Kastner, T., Plutzar, C., Bais, A. L. S., Carval-
hais, N., Fetzel, T., Gingrich, S., Haberl, H., Lauk, C.,
Niedertscheider, M., Pongratz, J., Thurner, M., and Luys-
saert, S.: Unexpectedly large impact of forest management
and grazing on global vegetation biomass, Nature, 553, 73–76,
https://doi.org/10.1038/nature25138, 2018.
Estes, L., Chen, P., Debats, S., Evans, T., Ferreira, S., Kuemmerle,
T., Ragazzo, G., Sheffield, J., Wolf, A., Wood, E., and Caylor,
K.: A large-area, spatially continuous assessment of land cover
map error and its impact on downstream analyses, Glob. Change
Biol., 24, 322–337, https://doi.org/10.1111/gcb.13904, 2018.
FAO: Transforming food and agriculture to achieve the SDGs, FAO
UN, Rome, Italy, 2018.
Fischer, G., Nachtergaele, F. O., Prieler, S., Teixeira, E., Tóth, G.,
Velthuizen, H. V., Verelst, L., and Wiberg, D.: Global Agro-
ecological Zones (GAEZ v3.0), IIASA, Laxenburg, Austria and
FAO, Rome, Italy, 2012.
Fritz, S., McCallum, I., Schill, C., Perger, C., See, L., Schep-
aschenko, D., van der Velde, M., Kraxner, F., and Ober-
steiner, M.: Geo-Wiki: An online platform for improving
global land cover, Environ. Modell. Softw., 31, 110–123,
https://doi.org/10.1016/j.envsoft.2011.11.015, 2012.
Fu, B. and Liu, Y.: The theories and methods for systematically
understanding land resource, Chinese Sci. Bull., 64, 2172–2179,
2019.
Gong, P., Liu, H., Zhang, M., Li, C., Wang, J., Huang, H., Clinton,
N., Ji, L., Li, W., Bai, Y., Chen, B., Xu, B., Zhu, Z., Yuan, C.,
Ping Suen, H., Guo, J., Xu, N., Li, W., Zhao, Y., Yang, J., Yu, C.,
Wang, X., Fu, H., Yu, L., Dronova, I., Hui, F., Cheng, X., Shi, X.,
Xiao, F., Liu, Q., and Song, L.: Stable classification with limited
sample: transferring a 30-m resolution sample set collected in
2015 to mapping 10-m resolution global land cover in 2017, Sci.
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Page 27
Q. Yu et al.: A cultivated planet in 2010 – Part 2
3571
Bull., 64, 370–373, https://doi.org/10.1016/j.scib.2019.03.002,
2019.
Herrmann, A.: Biogas Production from Maize: Current State, Chal-
lenges and Prospects, 2, Agronomic and Environmental Aspects,
BioEnerg. Res., 6, 372–387, https://doi.org/10.1007/s12155-
012-9227-x, 2013.
IFPRI: Global Spatially-Disaggregated Crop Production Statis-
tics Data for 2010 Version 2.0, Harvard Dataverse, V4,
https://doi.org/10.7910/DVN/PRFF8V, 2019.
Iizumi, T. and Sakai, T.: The global dataset of histori-
cal yields for major crops 1981–2016, Sci. Data, 7, 97,
https://doi.org/10.1038/s41597-020-0433-7, 2020.
Joglekar,
A.
K.
B.,
Wood-Sichra,
U.,
and
Pardey,
P.
G.:
Pixelating
crop
production:
Consequences
of
methodological
choices,
PloS
one,
14,
e0212281,
https://doi.org/10.1371/journal.pone.0212281, 2019.
Koks, E. E., Rozenberg, J., Zorn, C., Tariverdi, M., Vousdoukas,
M., Fraser, S. A., Hall, J. W., and Hallegatte, S.: A global multi-
hazard risk analysis of road and railway infrastructure assets,
Nat. Commun., 10, 2677, https://doi.org/10.1038/s41467-019-
10442-3, 2019.
Leff, B., Ramankutty, N., and Foley, J. A.: Geographic distribution
of major crops across the world, Global Biogeochem. Cy., 18,
GB1009, https://doi.org/10.1029/2003gb002108, 2004.
Lev-Yadun,
S.,
Gopher,
A.,
and
Abbo,
S.:
The
Cra-
dle
of
Agriculture,
Science,
288,
1602–1603,
https://doi.org/10.1126/science.288.5471.1602, 2000.
Li, Z., Tan, J., Tang, P., Chen, H., Zhang, L., Liu, H., Wu, W., Tang,
H., Yang, P., and Liu, Z.: Spatial distribution of maize in response
to climate change in northeast China during 1980–2010, J. Ge-
ogr. Sci., 26, 3–14, https://doi.org/10.1007/s11442-016-1250-y,
2016.
Liu, Z., Li, Z., Tang, P., Li, Z., Wu, W., Yang, P., You, L., and
Tang, H.: Change analysis of rice area and production in China
during the past three decades, J. Geogr. Sci., 23, 1005–1018,
https://doi.org/10.1007/s11442-013-1059-x, 2013.
Losch, B., Freguin-Gresh, S., and White, E. T.: Structural transfor-
mation and rural change revisited: challenges for late develop-
ing countries in a globalizing world, African Development Fo-
rum Series, Washington, DC, USA, https://doi.org/10.1596/978-
0-8213-9512-7, 2012.
Lu, M., Wu, W., You, L., See, L., Fritz, S., Yu, Q., Wei, Y., Chen, D.,
Yang, P., and Xue, B.: A cultivated planet in 2010 – Part 1: The
global synergy cropland map, Earth Syst. Sci. Data, 12, 1913–
1928, https://doi.org/10.5194/essd-12-1913-2020, 2020.
Monfreda, C., Ramankutty, N., and Foley, J. A.: Farm-
ing the planet: 2. Geographic distribution of crop ar-
eas, yields, physiological types, and net primary production
in the year 2000, Global Biogeochem. Cy., 22, GB1022,
https://doi.org/10.1029/2007gb002947, 2008.
Nanni, A. S., Sloan, S., Aide, T. M., Graesser, J., Edwards,
D., and Grau, H. R.: The neotropical reforestation hotspots:
A biophysical and socioeconomic typology of contempo-
rary forest expansion, Global Environ. Chang., 54, 148–159,
https://doi.org/10.1016/j.gloenvcha.2018.12.001, 2019.
Ornetsmüller, C., Castella, J.-C., Thanichanon, P., Lestrelin, G., and
Verburg, P. H.: Modelling the location and spatial pattern of a
crop boom. A case study from Laos, Environ. Sci. Policy, 99,
58–71, https://doi.org/10.1016/j.envsci.2019.04.001, 2019.
Peters, M. K., Hemp, A., Appelhans, T., Becker, J. N., Behler, C.,
Classen, A., Detsch, F., Ensslin, A., Ferger, S. W., Frederiksen,
S. B., Gebert, F., Gerschlauer, F., Gütlein, A., Helbig-Bonitz, M.,
Hemp, C., Kindeketa, W. J., Kühnel, A., Mayr, A. V., Mwan-
gomo, E., Ngereza, C., Njovu, H. K., Otte, I., Pabst, H., Ren-
ner, M., Röder, J., Rutten, G., Schellenberger Costa, D., Sierra-
Cornejo, N., Vollstädt, M. G. R., Dulle, H. I., Eardley, C. D.,
Howell, K. M., Keller, A., Peters, R. S., Ssymank, A., Kakengi,
V., Zhang, J., Bogner, C., Böhning-Gaese, K., Brandl, R., Her-
tel, D., Huwe, B., Kiese, R., Kleyer, M., Kuzyakov, Y., Nauss,
T., Schleuning, M., Tschapka, M., Fischer, M., and Steffan-
Dewenter, I.: Climate-land-use interactions shape tropical moun-
tain biodiversity and ecosystem functions, Nature, 568, 88–92,
https://doi.org/10.1038/s41586-019-1048-z, 2019.
Poore, J. and Nemecek, T.: Reducing food’s environmental im-
pacts through producers and consumers, Science, 360, 987–992,
https://doi.org/10.1126/science.aaq0216, 2018.
Portmann, F. T., Siebert, S., and Döll, P.: MIRCA2000 – Global
monthly irrigated and rainfed crop areas around the year
2000: A new high-resolution data set for agricultural and hy-
drological modeling, Global Biogeochem. Cy., 24, GB1011,
https://doi.org/10.1029/2008gb003435, 2010.
Porwollik, V., Rolinski, S., Heinke, J., and Müller, C.: Generating
a rule-based global gridded tillage dataset, Earth Syst. Sci. Data,
11, 823–843, https://doi.org/10.5194/essd-11-823-2019, 2019.
Prestele, R., Hirsch, A. L., Davin, E. L., Seneviratne, S.
I., and Verburg, P. H.: A spatially explicit representa-
tion of conservation agriculture for application in global
change
studies,
Glob.
Change
Biol.,
24,
4038–4053,
https://doi.org/10.1111/gcb.14307, 2018.
Pretty,
J.:
Intensification
for
redesigned
and
sustain-
able
agricultural
systems,
Science,
362,
eaav0294,
https://doi.org/10.1126/science.aav0294, 2018.
Searchinger, T. D., Wirsenius, S., Beringer, T., and Du-
mas, P.: Assessing the efficiency of changes in land use
for
mitigating
climate
change,
Nature,
564,
249–253,
https://doi.org/10.1038/s41586-018-0757-z, 2018.
Siebert, S. and Döll, P.: Quantifying blue and green virtual wa-
ter contents in global crop production as well as potential pro-
duction losses without irrigation, J. Hydrol., 384, 198–217,
https://doi.org/10.1016/j.jhydrol.2009.07.031, 2010.
Siebert, S., Henrich, V., Frenken, K., and Burke, J.: Update of
the digital global map of irrigation areas to version 5, Rheinis-
che Friedrich-Wilhelms Universität, Bonn, Germany and FAO,
Rome, Italy, 2013.
Sun, J., Mooney, H., Wu, W., Tang, H., Tong, Y., Xu, Z., Huang,
B., Cheng, Y., Yang, X., Wei, D., Zhang, F., and Liu, J.: Im-
porting food damages domestic environment: Evidence from
global soybean trade, P. Natl. Acad. Sci. USA, 115, 5415–5419,
https://doi.org/10.1073/pnas.1718153115, 2018.
Verburg, P. H., Neumann, K., and Nol, L.: Challenges in us-
ing land use and land cover data for global change studies,
Glob. Change Biol., 17, 974–989, https://doi.org/10.1111/j.1365-
2486.2010.02307.x, 2011.
Wood-Sichra, U., Joglekar, A. B., and You, L.: Spatial Produc-
tion Allocation Model (SPAM) 2005: Technical Documentation,
International Food Policy Research Institute (IFPRI), Washing-
ton, D.C., and International Science and Technology Practice
https://doi.org/10.5194/essd-12-3545-2020
Earth Syst. Sci. Data, 12, 3545–3572, 2020
Page 28
3572
Q. Yu et al.: A cultivated planet in 2010 – Part 2
and Policy (InSTePP) Center, University of Minnesota, St. Paul,
2016.
Xie, W., Xiong, W., Pan, J., Ali, T., Cui, Q., Guan, D., Meng, J.,
Mueller, N. D., Lin, E., and Davis, S. J.: Decreases in global beer
supply due to extreme drought and heat, Nat. Plants, 4, 964–973,
https://doi.org/10.1038/s41477-018-0263-1, 2018.
You, L. and Wood, S.: An entropy approach to spatial disag-
gregation of agricultural production, Agr. Syst., 90, 329–347,
https://doi.org/10.1016/j.agsy.2006.01.008, 2006.
You, L., Wood, S., and Wood-Sichra, U.: Generating plausible crop
distribution maps for Sub-Saharan Africa using a spatially disag-
gregated data fusion and optimization approach, Agr. Syst., 99,
126–140, https://doi.org/10.1016/j.agsy.2008.11.003, 2009.
You, L., Wood, S., Wood-Sichra, U., and Wu, W.: Generating global
crop distribution maps: From census to grid, Agr. Syst., 127, 53–
60, https://doi.org/10.1016/j.agsy.2014.01.002, 2014.
Yu, Q., Wu, W., Yang, P., Li, Z., Xiong, W., and Tang, H.: Proposing
an interdisciplinary and cross-scale framework for global change
and food security researches, Agr. Ecosyst. Environ., 156, 57–71,
https://doi.org/10.1016/j.agee.2012.04.026, 2012.
Yu, Q., Shi, Y., Tang, H., Yang, P., Xie, A., Liu, B., and Wu, W.:
eFarm: A tool for better observing agricultural land systems,
Sensors, 17, 453, https://doi.org/10.3390/s17030453, 2017a.
Yu, Q., Wu, W., You, L., Zhu, T., van Vliet, J., Verburg, P. H.,
Liu, Z., Li, Z., Yang, P., Zhou, Q., and Tang, H.: Assessing
the harvested area gap in China, Agr. Syst., 153, 212–220,
https://doi.org/10.1016/j.agsy.2017.02.003, 2017b.
Zhang, G., Xiao, X., Biradar, C. M., Dong, J., Qin, Y.,
Menarguez, M. A., Zhou, Y., Zhang, Y., Jin, C., Wang,
J., Doughty, R. B., Ding, M., and Moore, B.: Spatiotem-
poral patterns of paddy rice croplands in China and In-
dia from 2000 to 2015, Sci. Total Environ., 579, 82–92,
https://doi.org/10.1016/j.scitotenv.2016.10.223, 2017.
Earth Syst. Sci. Data, 12, 3545–3572, 2020
https://doi.org/10.5194/essd-12-3545-2020
Document Outline