Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model
Identifieur interne : 001312 ( Istex/Corpus ); précédent : 001311; suivant : 001313Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model
Auteurs : Liangzhi You ; Stanley Wood ; Ulrike Wood-Sichra ; Jordan ChamberlinSource :
- Information Development [ 0266-6669 ] ; 2007-05.
Abstract
Agricultural production statistics are fundamental parameters for agriculture policy research. Information on acreage and yields of important crops is critical for understanding trends within what is the most important economic sector of many developing countries. Sub-national data — i.e. data organized by administrative units such as regions or districts — enable the analysis of patterns within countries that may highlight important policy issues, such as the need to allocate resources to underproductive areas. However, collecting sub-national data is difficult for developing countries with limited resources. Even with great effort, and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill these spatial data gaps we developed a model to disaggregate production data from coarser to finer spatial units. Using a cross-entropy approach, our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. In addition to more detailed information, the organization of production information in geographic grids allows for greater analytical possibilities through geographic information systems. The allocation model works on the basis of available evidence of mapped indicators of agricultural production, which include farming systems, land cover, crop biophysical suitability surfaces, commodity prices and local market access. This article describes the generation of crop distribution maps for Sub-Saharan Africa for the year 2000 using the spatial allocation model and discusses the importance of such maps for development analysis and planning.
Url:
DOI: 10.1177/0266666907078670
Links to Exploration step
ISTEX:7B56D53FD66612A3E586489E71BA0D2687E53E51Le document en format XML
<record><TEI wicri:istexFullTextTei="biblStruct"><teiHeader><fileDesc><titleStmt><title xml:lang="en">Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</title><author><name sortKey="You, Liangzhi" sort="You, Liangzhi" uniqKey="You L" first="Liangzhi" last="You">Liangzhi You</name><affiliation><mods:affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</mods:affiliation></affiliation></author><author><name sortKey="Wood, Stanley" sort="Wood, Stanley" uniqKey="Wood S" first="Stanley" last="Wood">Stanley Wood</name><affiliation><mods:affiliation>International Food Policy Research Institute</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute</mods:affiliation></affiliation></author><author><name sortKey="Wood Sichra, Ulrike" sort="Wood Sichra, Ulrike" uniqKey="Wood Sichra U" first="Ulrike" last="Wood-Sichra">Ulrike Wood-Sichra</name><affiliation><mods:affiliation>International Food Policy Research Institute in Washington, DC</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute in Washington, DC</mods:affiliation></affiliation></author><author><name sortKey="Chamberlin, Jordan" sort="Chamberlin, Jordan" uniqKey="Chamberlin J" first="Jordan" last="Chamberlin">Jordan Chamberlin</name><affiliation><mods:affiliation></mods:affiliation></affiliation><affiliation><mods:affiliation>E-mail: j.chamberlin@cgiar.org</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute, j.chamberlin@cgiar.org</mods:affiliation></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">ISTEX</idno><idno type="RBID">ISTEX:7B56D53FD66612A3E586489E71BA0D2687E53E51</idno><date when="2007" year="2007">2007</date><idno type="doi">10.1177/0266666907078670</idno><idno type="url">https://api.istex.fr/document/7B56D53FD66612A3E586489E71BA0D2687E53E51/fulltext/pdf</idno><idno type="wicri:Area/Istex/Corpus">001312</idno><idno type="wicri:explorRef" wicri:stream="Istex" wicri:step="Corpus" wicri:corpus="ISTEX">001312</idno></publicationStmt><sourceDesc><biblStruct><analytic><title level="a" type="main" xml:lang="en">Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</title><author><name sortKey="You, Liangzhi" sort="You, Liangzhi" uniqKey="You L" first="Liangzhi" last="You">Liangzhi You</name><affiliation><mods:affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</mods:affiliation></affiliation></author><author><name sortKey="Wood, Stanley" sort="Wood, Stanley" uniqKey="Wood S" first="Stanley" last="Wood">Stanley Wood</name><affiliation><mods:affiliation>International Food Policy Research Institute</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute</mods:affiliation></affiliation></author><author><name sortKey="Wood Sichra, Ulrike" sort="Wood Sichra, Ulrike" uniqKey="Wood Sichra U" first="Ulrike" last="Wood-Sichra">Ulrike Wood-Sichra</name><affiliation><mods:affiliation>International Food Policy Research Institute in Washington, DC</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute in Washington, DC</mods:affiliation></affiliation></author><author><name sortKey="Chamberlin, Jordan" sort="Chamberlin, Jordan" uniqKey="Chamberlin J" first="Jordan" last="Chamberlin">Jordan Chamberlin</name><affiliation><mods:affiliation></mods:affiliation></affiliation><affiliation><mods:affiliation>E-mail: j.chamberlin@cgiar.org</mods:affiliation></affiliation><affiliation><mods:affiliation>International Food Policy Research Institute, j.chamberlin@cgiar.org</mods:affiliation></affiliation></author></analytic><monogr></monogr><series><title level="j">Information Development</title><idno type="ISSN">0266-6669</idno><idno type="eISSN">1741-6469</idno><imprint><publisher>Sage Publications</publisher><pubPlace>Sage UK: London, England</pubPlace><date type="published" when="2007-05">2007-05</date><biblScope unit="volume">23</biblScope><biblScope unit="issue">2-3</biblScope><biblScope unit="page" from="151">151</biblScope><biblScope unit="page" to="159">159</biblScope></imprint><idno type="ISSN">0266-6669</idno></series><idno type="istex">7B56D53FD66612A3E586489E71BA0D2687E53E51</idno><idno type="DOI">10.1177/0266666907078670</idno><idno type="ArticleID">10.1177_0266666907078670</idno></biblStruct></sourceDesc><seriesStmt><idno type="ISSN">0266-6669</idno></seriesStmt></fileDesc><profileDesc><textClass></textClass><langUsage><language ident="en">en</language></langUsage></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">Agricultural production statistics are fundamental parameters for agriculture policy research. Information on acreage and yields of important crops is critical for understanding trends within what is the most important economic sector of many developing countries. Sub-national data — i.e. data organized by administrative units such as regions or districts — enable the analysis of patterns within countries that may highlight important policy issues, such as the need to allocate resources to underproductive areas. However, collecting sub-national data is difficult for developing countries with limited resources. Even with great effort, and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill these spatial data gaps we developed a model to disaggregate production data from coarser to finer spatial units. Using a cross-entropy approach, our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. In addition to more detailed information, the organization of production information in geographic grids allows for greater analytical possibilities through geographic information systems. The allocation model works on the basis of available evidence of mapped indicators of agricultural production, which include farming systems, land cover, crop biophysical suitability surfaces, commodity prices and local market access. This article describes the generation of crop distribution maps for Sub-Saharan Africa for the year 2000 using the spatial allocation model and discusses the importance of such maps for development analysis and planning.</div></front></TEI><istex><corpusName>sage</corpusName><author><json:item><name>Liangzhi You</name><affiliations><json:string>International Food Policy Research Institute (IFPRI) in Washington, DC</json:string><json:string>International Food Policy Research Institute (IFPRI) in Washington, DC</json:string></affiliations></json:item><json:item><name>Stanley Wood</name><affiliations><json:string>International Food Policy Research Institute</json:string><json:string>International Food Policy Research Institute</json:string></affiliations></json:item><json:item><name>Ulrike Wood-Sichra</name><affiliations><json:string>International Food Policy Research Institute in Washington, DC</json:string><json:string>International Food Policy Research Institute in Washington, DC</json:string></affiliations></json:item><json:item><name>Jordan Chamberlin</name><affiliations><json:null></json:null><json:string>E-mail: j.chamberlin@cgiar.org</json:string><json:string>International Food Policy Research Institute, j.chamberlin@cgiar.org</json:string></affiliations></json:item></author><subject><json:item><lang><json:string>eng</json:string></lang><value>cross entropy</value></json:item><json:item><lang><json:string>eng</json:string></lang><value>spatial allocation</value></json:item><json:item><lang><json:string>eng</json:string></lang><value>agricultural production</value></json:item><json:item><lang><json:string>eng</json:string></lang><value>crop suitability</value></json:item><json:item><lang><json:string>eng</json:string></lang><value>geographic information systems</value></json:item><json:item><lang><json:string>eng</json:string></lang><value>Sub-Saharan Africa</value></json:item></subject><articleId><json:string>10.1177_0266666907078670</json:string></articleId><language><json:string>eng</json:string></language><originalGenre><json:string>research-article</json:string></originalGenre><abstract>Agricultural production statistics are fundamental parameters for agriculture policy research. Information on acreage and yields of important crops is critical for understanding trends within what is the most important economic sector of many developing countries. Sub-national data — i.e. data organized by administrative units such as regions or districts — enable the analysis of patterns within countries that may highlight important policy issues, such as the need to allocate resources to underproductive areas. However, collecting sub-national data is difficult for developing countries with limited resources. Even with great effort, and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill these spatial data gaps we developed a model to disaggregate production data from coarser to finer spatial units. Using a cross-entropy approach, our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. In addition to more detailed information, the organization of production information in geographic grids allows for greater analytical possibilities through geographic information systems. The allocation model works on the basis of available evidence of mapped indicators of agricultural production, which include farming systems, land cover, crop biophysical suitability surfaces, commodity prices and local market access. This article describes the generation of crop distribution maps for Sub-Saharan Africa for the year 2000 using the spatial allocation model and discusses the importance of such maps for development analysis and planning.</abstract><qualityIndicators><score>8.578</score><pdfVersion>1.5</pdfVersion><pdfPageSize>595.276 x 841.89 pts (A4)</pdfPageSize><refBibsNative>true</refBibsNative><abstractCharCount>2042</abstractCharCount><pdfWordCount>4578</pdfWordCount><pdfCharCount>31413</pdfCharCount><pdfPageCount>9</pdfPageCount><abstractWordCount>293</abstractWordCount></qualityIndicators><title>Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</title><genre><json:string>research-article</json:string></genre><host><volume>23</volume><publisherId><json:string>IDV</json:string></publisherId><pages><last>159</last><first>151</first></pages><issn><json:string>0266-6669</json:string></issn><issue>2-3</issue><genre><json:string>journal</json:string></genre><language><json:string>unknown</json:string></language><eissn><json:string>1741-6469</json:string></eissn><title>Information Development</title></host><categories><wos><json:string>social science</json:string><json:string>information science & library science</json:string></wos><scienceMetrix></scienceMetrix></categories><publicationDate>2007</publicationDate><copyrightDate>2007</copyrightDate><doi><json:string>10.1177/0266666907078670</json:string></doi><id>7B56D53FD66612A3E586489E71BA0D2687E53E51</id><score>0.04863176</score><fulltext><json:item><extension>pdf</extension><original>true</original><mimetype>application/pdf</mimetype><uri>https://api.istex.fr/document/7B56D53FD66612A3E586489E71BA0D2687E53E51/fulltext/pdf</uri></json:item><json:item><extension>zip</extension><original>false</original><mimetype>application/zip</mimetype><uri>https://api.istex.fr/document/7B56D53FD66612A3E586489E71BA0D2687E53E51/fulltext/zip</uri></json:item><istex:fulltextTEI uri="https://api.istex.fr/document/7B56D53FD66612A3E586489E71BA0D2687E53E51/fulltext/tei"><teiHeader><fileDesc><titleStmt><title level="a" type="main" xml:lang="en">Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</title></titleStmt><publicationStmt><authority>ISTEX</authority><publisher>Sage Publications</publisher><pubPlace>Sage UK: London, England</pubPlace><availability><p>SAGE</p></availability><date>2007</date></publicationStmt><sourceDesc><biblStruct type="inbook"><analytic><title level="a" type="main" xml:lang="en">Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</title><author xml:id="author-1"><persName><forename type="first">Liangzhi</forename><surname>You</surname></persName><affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</affiliation><affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</affiliation></author><author xml:id="author-2"><persName><forename type="first">Stanley</forename><surname>Wood</surname></persName><affiliation>International Food Policy Research Institute</affiliation><affiliation>International Food Policy Research Institute</affiliation></author><author xml:id="author-3"><persName><forename type="first">Ulrike</forename><surname>Wood-Sichra</surname></persName><affiliation>International Food Policy Research Institute in Washington, DC</affiliation><affiliation>International Food Policy Research Institute in Washington, DC</affiliation></author><author xml:id="author-4"><persName><forename type="first">Jordan</forename><surname>Chamberlin</surname></persName><email>j.chamberlin@cgiar.org</email><affiliation></affiliation><affiliation>International Food Policy Research Institute, j.chamberlin@cgiar.org</affiliation></author></analytic><monogr><title level="j">Information Development</title><idno type="pISSN">0266-6669</idno><idno type="eISSN">1741-6469</idno><imprint><publisher>Sage Publications</publisher><pubPlace>Sage UK: London, England</pubPlace><date type="published" when="2007-05"></date><biblScope unit="volume">23</biblScope><biblScope unit="issue">2-3</biblScope><biblScope unit="page" from="151">151</biblScope><biblScope unit="page" to="159">159</biblScope></imprint></monogr><idno type="istex">7B56D53FD66612A3E586489E71BA0D2687E53E51</idno><idno type="DOI">10.1177/0266666907078670</idno><idno type="ArticleID">10.1177_0266666907078670</idno></biblStruct></sourceDesc></fileDesc><profileDesc><creation><date>2007</date></creation><langUsage><language ident="en">en</language></langUsage><abstract xml:lang="en"><p>Agricultural production statistics are fundamental parameters for agriculture policy research. Information on acreage and yields of important crops is critical for understanding trends within what is the most important economic sector of many developing countries. Sub-national data — i.e. data organized by administrative units such as regions or districts — enable the analysis of patterns within countries that may highlight important policy issues, such as the need to allocate resources to underproductive areas. However, collecting sub-national data is difficult for developing countries with limited resources. Even with great effort, and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill these spatial data gaps we developed a model to disaggregate production data from coarser to finer spatial units. Using a cross-entropy approach, our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. In addition to more detailed information, the organization of production information in geographic grids allows for greater analytical possibilities through geographic information systems. The allocation model works on the basis of available evidence of mapped indicators of agricultural production, which include farming systems, land cover, crop biophysical suitability surfaces, commodity prices and local market access. This article describes the generation of crop distribution maps for Sub-Saharan Africa for the year 2000 using the spatial allocation model and discusses the importance of such maps for development analysis and planning.</p></abstract><textClass><keywords scheme="keyword"><list><head>keywords</head><item><term>cross entropy</term></item><item><term>spatial allocation</term></item><item><term>agricultural production</term></item><item><term>crop suitability</term></item><item><term>geographic information systems</term></item><item><term>Sub-Saharan Africa</term></item></list></keywords></textClass></profileDesc><revisionDesc><change when="2007-05">Published</change></revisionDesc></teiHeader></istex:fulltextTEI><json:item><extension>txt</extension><original>false</original><mimetype>text/plain</mimetype><uri>https://api.istex.fr/document/7B56D53FD66612A3E586489E71BA0D2687E53E51/fulltext/txt</uri></json:item></fulltext><metadata><istex:metadataXml wicri:clean="corpus sage not found" wicri:toSee="no header"><istex:xmlDeclaration>version="1.0" encoding="UTF-8"</istex:xmlDeclaration><istex:docType PUBLIC="-//NLM//DTD Journal Publishing DTD v2.3 20070202//EN" URI="journalpublishing.dtd" name="istex:docType"></istex:docType><istex:document><article article-type="research-article" dtd-version="2.3" xml:lang="EN"><front><journal-meta><journal-id journal-id-type="hwp">spidv</journal-id><journal-id journal-id-type="publisher-id">IDV</journal-id><journal-title>Information Development</journal-title><issn pub-type="ppub">0266-6669</issn><publisher><publisher-name>Sage Publications</publisher-name><publisher-loc>Sage UK: London, England</publisher-loc></publisher></journal-meta><article-meta><article-id pub-id-type="doi">10.1177/0266666907078670</article-id><article-id pub-id-type="publisher-id">10.1177_0266666907078670</article-id><article-categories><subj-group subj-group-type="heading"><subject>Articles</subject></subj-group></article-categories><title-group><article-title>Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</article-title></title-group><contrib-group><contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>You</surname><given-names>Liangzhi</given-names></name><aff>International Food Policy Research Institute (IFPRI) in Washington, DC</aff></contrib><contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Wood</surname><given-names>Stanley</given-names></name><aff>International Food Policy Research Institute</aff></contrib><contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Wood-Sichra</surname><given-names>Ulrike</given-names></name><aff>International Food Policy Research Institute in Washington, DC</aff></contrib><contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Chamberlin</surname><given-names>Jordan</given-names></name><aff>International Food Policy Research Institute, <email xlink:type="simple">j.chamberlin@cgiar.org</email></aff></contrib></contrib-group><pub-date pub-type="ppub"><month>05</month><year>2007</year></pub-date><volume>23</volume><issue>2-3</issue><fpage>151</fpage><lpage>159</lpage><abstract><p><bold>Agricultural production statistics are fundamental parameters for agriculture policy research. Information on acreage and yields of important crops is critical for understanding trends within what is the most important economic sector of many developing countries. Sub-national data — i.e. data organized by administrative units such as regions or districts — enable the analysis of patterns within countries that may highlight important policy issues, such as the need to allocate resources to underproductive areas. However, collecting sub-national data is difficult for developing countries with limited resources. Even with great effort, and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill these spatial data gaps we developed a model to disaggregate production data from coarser to finer spatial units. Using a cross-entropy approach, our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. In addition to more detailed information, the organization of production information in geographic grids allows for greater analytical possibilities through geographic information systems. The allocation model works on the basis of available evidence of mapped indicators of agricultural production, which include farming systems, land cover, crop biophysical suitability surfaces, commodity prices and local market access. This article describes the generation of crop distribution maps for Sub-Saharan Africa for the year 2000 using the spatial allocation model and discusses the importance of such maps for development analysis and planning.</bold></p></abstract><kwd-group><kwd>cross entropy</kwd><kwd>spatial allocation</kwd><kwd>agricultural production</kwd><kwd>crop suitability</kwd><kwd>geographic information systems</kwd><kwd>Sub-Saharan Africa</kwd></kwd-group><custom-meta-wrap><custom-meta xlink:type="simple"><meta-name>sagemeta-type</meta-name><meta-value>Journal Article</meta-value></custom-meta><custom-meta xlink:type="simple"><meta-name>search-text</meta-name><meta-value>151 Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation ModelThe use of geographic information systems offers strong possibilities for understanding and manipulating geographical conditioners of agriculture through policy interventions. A spatial allocation model is an effective way of mapping detailed patterns of crop production from coarser input data SAGE Publications, Inc.200710.1177/0266666907078670 Liangzhi You International Food Policy Research Institute (IFPRI) in Washington, DC StanleyWood International Food Policy Research Institute UlrikeWood-Sichra International Food Policy Research Institute in Washington, DC JordanChamberlin International Food Policy Research Institute, j.chamberlin@cgiar.org Liangzhi You, Stanley Wood, Ulrike Wood-Sichra and Jordan Chamberlin THE NEED FOR MORE DETAILED INFORMATION ON THE LOCATION OF AGRICULTURAL PRODUCTION In the design and targeting of rural development strategies to stimulate economic growth and alleviate poverty in the developing world, the sustainable growth of agriculture is a priority. Because agricultural production choices and potentials are determined in part by location characteristics (such as water avail- ability or the presence of markets), the impact of devel- opment strategies depends, to a large extent, upon our better understanding of the geographical patterns of agricultural activities and performance (Wood et al., 1999). Because of pronounced geographical variation in production (across regions, countries and agro- ecological zones), spatial data1 on production provide extremely valuable guidance for rural development policy. Agricultural economists increasingly argue for the importance of detailed spatial data in research for rural development (Nelson, 2002; Staal et al., 2002; Luijten, 2002; Bell and Irwin, 2002; Anselin, 2002). Spatially disaggregated agricultural production statistics are required for distinguishing patterns or trends which are heterogeneous within countries. However, collecting very detailed sub-national data (e.g. at the county or parish level) is difficult for developing countries with limited resources. Even with great effort and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill such data gaps we developed a model to dis- aggregate production data from coarser to finer spatial units (You and Wood, 2003). Our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. The allocation model works on the basis of inferring likely production locations from multiple indicators compiled within a geographic information system. Available geographical indicators include maps of farming systems, land cover, crop- specific biophysical suitability, commodity prices and local market access. The model employs a cross-entropy approach to manage inputs with different levels of likelihood in indicating the specific locations of agri- cultural production. Previous validation of this model showed that production data disaggregated from coarse reporting units agrees tolerably well with available data from much smaller reporting units (You and Wood, 2006). This paper describes the application of this model to generate crop distribution maps for Sub-Saharan Africa (SSA) for the year 2000. In the following sections, we first present different mapped indicators of production location, then describe how the spatial allocation model uses such inputs to produce detailed production maps from statistical data. We illustrate this with several crop distribution maps for SSA. We conclude with some remarks on the value of such datasets for guiding rural development. AGRICULTURAL HETEROGENEITY AND PRODUCTION INFORMATION The total agricultural land in SSA was about 903 million hectares in 2000, which represents about 40 percent of the land surface and clearly establishes agriculture as the most important land use in SSA. Most agricultural land, about 82 percent, is permanent pasture while the total 152 land under annual and permanent crops is only about 161 million hectares (FAOSTAT, 2005). Most of the region's agriculture takes place in humid or sub-humid tropic zones with heterogeneous land cover, such as mixed forest, pasture, permanent and annual crop land. The majority of farmers in the region are smallholders and subsistence-oriented. Many staple crops, such as plantain and cassava, are commonly planted in household gardens and intermixed with other crops. Such complex production systems and mixed cropping patterns result in significant heterogeneity of local production patterns. Country-level production data are available from FAO (FAOSTAT, 2005). Despite great efforts to gather all available information, sub-national data2 coverage for our base year is quite spotty. Figure 1 shows the sub-national data coverage for the 20 selected crops. Just a few countries (Benin, Cameroon, Democratic Republic of Congo, Gambia, Guinea Bissau, Mozambique, Uganda and Zambia) have sub- national statistics available for nine or more crops. Quite a few countries (Angola, Republic of Congo, Gabon and Ivory Coast) have no sub-national data available at all. Some crops are better covered than others: production data for cowpeas, bean, maize and cassava are available for about 70 percent of all first- level sub-national units. Overall, only about 40 percent of production data for our 20 commodities are available at some sub-national level. As indicated in the introduction, the objective of the work described in this article is the disaggregation of these production statistics using a variety of spati- ally explicit indicators of the likelihood of local pro- duction. The following sub-sections describe the kinds of information used by the model to perform the allocation. Disaggregating Statistics by Production System External inputs such as irrigation, fertilizer, pesticide, affect agricultural production in many ways. Therefore, disaggregating the crop area into different production systems according to input level should improve the spatial allocation. In addition, agro-ecological suitabil- ity is also defined according to different input levels. For those area statistics which we have, either at country level or at sub-national level, we partition each crop prod- uction into four levels: irrigated, rainfed/high-input, rainfed/low-input and subsistence. Although subsistence cropsn SSA are almost all grown underow-input rain-fed conditions, we break reported crop areas into all four production systems through a mixture of guiding fac- tors: aggregate data and informal studies on technology use, as well as expert opinion. This process is described more fully in You et al. (2007). Population Density Profit relations are among the main determinants of the type and volume of agricultural production activ- ities and play a fundamental role in formulation of development plans and related decisions. For traded crops, farmers' decision what to grow and where depends very much on the profitability of the crop production. Even for subsistence crops the values of crop productions are also important factors for far- mers to make planting decisions. Markets, even as imperfect as they are in SSA (Kherallah et al., 2000), are an important influence on agricultural production systems. Market access is an indicator of transaction costs, based on proximity to the nearby markets. Market access affects both cost of production (via input market such as fertilizer, pesticide, seed, etc.) and gross revenue (via transportation cost and other transactions, for example). Currently we have no reliable data for road network and market distribution in SSA, and so we simply use population density as a proxy for market access. Figure 1. Sub-national data coverage map. Colour versions of gures are available in the online version, see Editorial for details. 153 Figure 2. Population density of Sub-Saharan Africa. We use a gridded set of estimates of population den- sities (Centre for International Earth Science Network, International Food Policy Research Institute and World Resources Institute, 2000). National figures have been reconciled to be consistent with United Nations population estimates for those years. Figure 2 is the map of population density for Sub-Saharan Africa. In the current model, population density provides spatial variability for crop prices, with higher densities representing potentially higher prices. On the other hand, the price variation with population depends on specific crops. For example the price of perishable products such as vegetables drops faster away from population centers than that of non-perishable products such as maize and rice. Land Cover Satellite-based land cover is an important input into the allocation model. As more and better remotely-sensed data become available with the new technology and the improvement of our ability to interpret remotely- sensed imagery, land cover data will dramatically im- prove the accuracy of our production allocation. Africa Land Cover 2000 from the Global Land Cover 2000 project (http://www-gvm.jrc.it/glc2000/) is used here. This is a regionally homogenous land cover dataset with relatively good representation of crop land. Figure 3 shows the cropland extent by pixel (5 × 5 minutes) from the above land cover dataset. Figure 3. Cropland extent for Sub-Saharan Africa. Irrigation Maps We use a global irrigation map which shows the amount of area equipped for irrigation as a percentage of the total area of a 9000 hectare grid cell3 (Siebert, Döll and Hoogeveen, 2001). We calculate the irrigated area by grid cell using this global irrigation map. Figure 4 shows irrigated area, by grid cell, for the region. In SSA, rainfed agriculture dominates, and irrigation is quite sparse (Figure 4). Madagascar and South Africa are the only two countries with substantial irrigated Figure 4. Irrigation map for Sub-Saharan Africa. 154 areas, about 1.43 million hectares and 1.15 million hectares respectively. Most countries in the region have very limited irrigation (e.g. Chad, D.R. Congo, Mauritania, Namibia). Agroclimatic Crop Suitability Different crops have different thermal, moisture and soil requirements for adequate growth, particularly under rainfed (i.e. non-irrigated) conditions. The agro-ecological zone (AEZ) methodology enables rational land-use planning on the basis of an inventory of land resources and evaluation of biophysical limitations and potentials (Fischer et al., 2001). This methodology provides a standardized framework for the characterization of climate, soil and terrain conditions relevant to agricultural production. Crop modeling and environmental matching procedures are used to identify crop-specific limitations of prevailing climate, soil and terrain resources, under assumed levels of inputs and management conditions. The AEZ methodology indicates maximum agronomically attainable crop yields and suitable crop areas for basic land resources units, rendered as grid-cells in the most recent digital databases (ibid.). From datasets elaborated from this methodology, we consider the mapped yield potentials for different crops under three basic production system types: irrigated/high-input, rainfed/high-input, and rainfed/low-input. For each crop, by each of these three input levels, we define suitable production land as the sum of the following four classes: very suitable, suitable, moderately suit- able and marginally suitable. Correspondingly, the yield is calculated as the area-weighted average of the above four suitable classes. (Food and Agriculture Organization of the United Nations, 1981; 2003). These three types correspond to the three production patterns defined before. The fourth one, subsistence, is always low-input rainfed production. Some crops have many types, such as highland and lowland maize germplasm, sub-divided by maturity class. The single `maize' crop surface is a composite in which each pixel would use the best variety. As an example, Figure 5 shows suit- ability surfaces for maize in SSA4. The maps show the distribution of suitable areas for maize grown under high input conditions, and the distribution of potential maize yields under irrigated conditions. As we can see, suitable conditions for maize production are found widely throughout the region, the major exceptions being the extremely northern countries such as Chad, Mali, Mauritania, Niger and the extreme southern part of the African continent. THE SPATIAL ALLOCATION MODEL As shown in the preceding sections, while there is a considerable amount of mapped information on the conditioning factors for agricultural production, the available statistical data on production in the region is compiled at very coarse levels. The purpose of the exer- cise we describe in this article is to use the more detailed set of indicators to disaggregate the statistical data. Figure 5. Suitability surfaces of maize. 155 For this to be a meaningful exercise, the logic underlying our approach to disaggregation (which we term `spatial allocation') must be plausible. Our spatial allocation model uses the concept of information entropy to sift through the various inputs (Jaynes, 1979; Golan, Judge and Miller, 1996). The crux of this approach is that it allows for variation in the rules which are used to interpret the importance of different inputs in inferring the distribution of prod- uction. This is important because of the different levels of information available across the areas over which the location production is to be estimated. The way the entropy approach works is by using inferential infor- mation where, and to the extent that, they are available and help to solve the allocation problem.5 We built our spatial allocation model on a cross- entropy approach6 which is modified to accommodate different levels of statistical data, i.e. a combination of national and different levels of sub-national data (You, Wood and Wood-Sichra, 2007). We first allocate the crop area and production into pixels by the best available production statistics and corresponding biophysical suitability maps. Then we input this preliminary allocation into an optimization model where the objective function is to minimize the differences between this allocation and a final allocation which reflects the influence of all the input data used to infer production locations. The process is as follows. First we constrain the sum of all allocated areas within those sub-national units with existing statistical data to be equal to the corresponding sub-national statistics. Secondly, the actual agricultural area in each pixel from satellite image is the upper limit for the area to be allocated to all crops. Thirdly, the model specifies that the allocated crop area cannot exceed what is suitable for the particular crop. Fourthly, the sum of all allocated irrigated areas in any pixel must not exceed the area equipped for irrigation indicated in the African map of irrigation. With all these constraints, our allocated areas would be consistent with the input information and the optimization process ensures the closest estimate to the initial, suitability-based allocation.7 The allocation method described here faces two major challenges. The first is the inconsistency among the various constraints due to imperfect data. For example, the sum of all the statistical crop areas may be even larger than the cropland available from satellite image at either national or sub-national levels. These inconsistencies happen for most countries in SSA. Our basic approach to this was to treat the reported production/area statistics as the `truth',8 and then modify the areas from other sources. For example, we scale up cropland areas from the land cover map if they are less than the sum of statistical crop areas available for a given region. Likewise, for those pixels where there is zero cropland but positive irrigated areas, we set the cropland areas equal to the irrigated areas. Such inconsistencies are generally minor, but do indicate some of the complexity of the allocation exercise. We run the model country by country for all coun- tries in Sub-Saharan Africa.9 In our current allocation, we use the year 2000 as our base year, and all statistical data are based on a three-year average (1999–2001). The model allocates production statistics for 20 crops: barley, beans, cassava, cocoa, coffee, cotton, cow peas, groundnuts, maize, millet, oil palm, plantain, potato, rice, sorghum, soybeans, sugar cane, sweet potato, wheat and yam. Altogether, these 20 crops cover more than 90 percent of total cropland in Sub-Saharan Africa and their total output is almost 40 percent of agricultural GDP in SSA. The model output is the estimated harvest area, production and yield of every pixel in a geographical grid of approximately 10 by 10 kilometer resolution. Within a geographic information system, such grids are directly mappable and amenable to further quanti- tative analysis. Figure 6 shows the crop area distribution maps for four staple crops: sorghum, maize, millet and rice. The full datasets for all 20 crops will be available online at IFPRI website (www.ifpri.org) soon. In addition, the input datasets described in the previous section are also available for free downloading at the above website. POTENTIAL APPLICATIONS AND FUTURE PLANS The spatial allocation model described here is an effective way of mapping relatively detailed patterns of crop production from much coarser input data. The approach utilizes information from various sources, starting from the best available production statistics, and guided by digital maps of land cover, biophysical crop suitability assessments, irrigation and population density, in order to generate plausible, disaggregated estimates of the distribution of crop production on the basis of a mapped grid of approximately 9000 hectare grid cells. Disaggregate data are useful for understanding the patterns of production. Understanding where trends take place is important for understanding why they 156 Figure 6. Estimated crop distribution maps of Sub-Saharan Africa. take place. Understanding the location of production relative to a country's infrastructure, services and other sectors will help to enable better planning, allocation of resources and targeting of interventions. Knowing the detailed patterns of productivity (i.e. yields) will help to diagnose and target underperforming areas. We present this model as an important step in achieving such spatially disaggregated information. Given the enormous diversity and site-specific nature of many African production systems – as well as their associated cultural, socioeconomic, and resource management issues – effective development strategies should take account of such spatial patterns. The characterization of geographical space by shared op- portunities and constraints to different development pathways is a fundamentally important framework for such strategy design (Wood et al., 1999; Omamo et al., 2006). Such opportunities may be biophysical (e.g. soil fertility endowment, climatic regime), infrastructural (e.g. availability of markets for agricultural production), or even social in nature (e.g., characteristics of farmers, capability of extension services). An important use of spatially disaggregate crop data (as from the model pre- sented here) would be to spatially define the agricultural 157 production systems and associated livelihood strategies (e.g. income from crop production) that form the basis of intervention strategies for meeting rural development objectives. The use of GIS to analyze the patterns of production and productivity offers strong possibilities for under- standing and manipulating geographical conditioners of agriculture through policy interventions. The fact that model output is also uniform in resolution and structure is important for its use within such systems. Such harmonized data are more easily introduced and manipulated within a GIS than the irregular structures of political or administrative boundaries in which statistics are usually compiled. In particular, data in a gridded format (i.e. raster-based spatial data) are flexibly compiled within a variety of analytical strata, significantly extending their uses. Consider the value of being able to calculate production shares by agroecological zones that span national boundaries – a task which is fairly impossible with traditional statistical data on agriculture, but readily performed with gridded production estimates 10. To be sure, further development of the model presented here still faces several challenges. First of all, better input (i.e. more disaggregated production statistics) will produce better output (i.e. more reliably disaggregated data and maps). As improvements in statistical reporting take place in developing countries in response to development of national statistical institutions, such data should be fed into updated model runs. Such time series output will also enable the analysis of spatially specific trends in production and productivity. Secondly, although the current model provides what appear, in the absence of `truth' regarding the real distribution of production, to be reasonable results, more work is under way to improve its performance. The obvious way forward is to improve the under- lying quality of the parameters currently included in the model, since the end results can only be as accurate as the input information. These include better approximations of the agricultural extent, more realistic crop suitability surfaces, and more research on the association between crop production and population density. On the other hand, we could also add more information into the model. For example, household or agricultural survey information on the location and quantity of crop production would provide a direct, sampled calibration of the entire crop distribution surface. If such information exists and it is of reasonable quality, it will definitely improve the estimation accuracy. We could also add some other behavioral assumptions. For example, it seems reasonable to assume that farmers would opt to plant a higher revenue crops in any given location, all other things being equal. But potential revenue is in reality a proxy for potential profitability, and some could argue that risk minimization might also play a role. Thus there are several options for further work in exploring alternative drivers of crop choice, both individually and in crop combinations, in each location. Such advances in our modeling approach, and con- sequent improvements in output maps, will continue to drive improved targeting of interventions to raise agricultural productivity. This is especially important for regions such as SSA, where large portions of the region's economies are driven by agricultural sectors with low inputs and poor productivity. Solving such problems is central to national and regional development objectives and will certainly be assisted by better and more disaggregate production data. Notes Agricultural production statistics are fundamental parameters for agriculture policy research. Information on acreage and yields of important crops is critical for understanding trends within what is the most important economic sector of many developing countries. Sub-national data — i.e. data organized by administrative units such as regions or districts — enable the analysis of patterns within countries that may highlight important policy issues, such as the need to allocate resources to underproductive areas. However, collecting sub-national data is difficult for developing countries with limited resources. Even with great effort, and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill these spatial data gaps we developed a model to disaggregate production data from coarser to finer spatial units. Using a cross-entropy approach, our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. In addition to more detailed information, the organization of production information in geographic grids allows for greater analytical possibilities through geographic information systems. The allocation model works on the basis of available evidence of mapped indicators of agricultural production, which include farming systems, land cover, crop biophysical suitability surfaces, commodity prices and local market access. This article describes the generation of crop distribution maps for Sub-Saharan Africa for the year 2000 using the spatial allocation model and discusses the importance of such maps for development analysis and planning. cross entropy spatial allocation agricultural production crop suitability geographic information systems Sub-Saharan Africa 1. Spatial data are sometimes referred to as geo-referenced data. Spatial data include some specific reference to location on the surface of the earth, expressed directly or indirectly through geographical coordinates, such as latitude and longitude. An important implied attribute of `spatial' data in economic studies is that of disaggregation to sub-national levels. 2. In this paper, `sub-national unit' refers to the first geopolitical division within a country, such as districts in Uganda, regions in Nigeria, or provinces in South Africa. Second level sub-national data are rarely available for SSA. 3. The model inputs and output described in this article all use a common grid resolution of 5 by 5 minutes of latitude and longitude. At the equator, this is equivalent to approximately 9.25 by 9.25 kilometers, or roughly 9000 hectares. All data used in our model are defined on the basis of the World Geodetic System 1984 Datum. 4. More details can be found under Global Agro-Ecological Zone 2000 at http://www.fao.org/ag/AGL/agll/gaez/ index.htm 5. A Wikipedia entry on cross entropy (http://en.wikepedia. org/wiki/Cross-entropy_method) may be helpful. 6. Shannon (1948) introduced information entropy to measure the uncertainty of the information in a given message, work which later gave rise to information theory. Information theory maintains that uncertainty is a statistical property of a message. Jaynes (1957) adopted the information entropy concept and proposed the maximum 158entropy principle in statistical inference: the least informative probability distribution can be found by maximizing the entropy of that distribution. In other words: in the absence of information to the contrary, all possible states of system are equally likely. Maximizing entropy is in fact a special case of minimizing cross-entropy with respect to a uniform distribution. The unique feature of the entropy approach is to overcome two empirical problems that hamper traditional econometrics: multi-collinearity and partially incomplete data (Golan, Judge and Miller, 1996). The idea is to remove irrelevant information at the beginning of a problem rather than taking pains to make dubious assumptions. Zellner (1988) described the advantage of the entropy approach as satisfying the `information conservation principle', by neither ignoring relevant input information nor injecting any false information. 7. Detailed model and equations could be found at You, Wood and Wood-Sichra (2007). 8. While we are not claiming that the reported agricultural statistics (either national or sub-national) are more accurate than, say, cropland indications derived from satellite imagery, we do establish the reported statistics as the basis for the ultimate crop maps. Thus, all the output grid cells corresponding to a given reporting unit (say, a province) will sum to that unit's reported total. 9. Except for some island countries such as Mayotte, Seychelles which have no or little agricultural production. 10. See Wood and Chamberlin, 2003 for fuller discussion of the value of spatially disaggregated agricultural data for rural development analysis and planning. References Anselin, L. (2002) Under the hood: issues in the specification andinterpretationofspatialregressionmodels . Agricultural Economics, 27, 247—267. Bell, K. and Irwin, E.G. (2002) Spatially explicit micro-level modeling of land use change at the rural-urban interface. Agricultural Economics , 27, 217—232. Center for International Earth Science Information Network (CIESIN) , Columbia University; International Food Policy Research Institute (IFPRI) and World Resources Institute (WRI). (2000) Gridded Population of the World (GPW), Version 2. Palisades, NY: CIESIN, Columbia University. Available at http://sedac.ciesin.columbia. edu/plue/gpw. Faostat. (2005) http://faostat.fao.org/default.aspx . Last accessed in October 2005. Fischer, Gunther, Shah, M., Velthuizen, H. and Nachtergaele, F. (2001) Global agro-ecological assessment for agriculture in the 21st century. International Institute for Applied Systems Analysis, Laxenburg, Austria. Food and Agriculture Organization of the United Nations (FAO). (1981) Report of the Agro-Ecological Zones Project. World Soil Resources Report, No 48, Vol.1—4, Rome, FAO. Food and Agriculture Organization of the United Nations (FAO). (2003) http://www.fao.org/ag/AGL/agll/gaez/ index.htm. Last accessed in July 2003. Golan, Amos; Judge, G. and Miller, D. (1996) Maximum entropy econometrics: robust estimation with limited data. New York: John Wiley & Sons. Jaynes, E.T. (1957) Information theory and statistical methods I. Physics Review, 106 (1957): 620—630. Jaynes, E.T. (1979) Where do we stand on maximum entropy? In The maximum entropy formalism. R. D. Levine and M. Tribus (eds.) MIT Press, Cambridge, MA, p.15. Kherallah, M., Delgado, C., Gabre-Madhin, E., Minot, N. and Johnson, M. (2000) The road half traveled: agricultural market reform in Sub-Sahara Africa. Food Policy Report. International Food Policy Research Institute, Washington, DC, USA. Luijten, J.C. (2002) A systematic method for generating land use patterns using stochastic rules and basic landscape characteristics: results from a Colombian hillside watershed. Agriculture Ecosystem & Environment, 95, 427—441. Nelson, G.C. (2002) Introduction to the special issue on spatial analysis for agricultural economists. Agricultural Economics, 27, 197—200. Omamo, S.W., Diao, X., Wood, S., Chamberlin, J., You, L., Benin, S., Wood-Sichra, U., Tatwangire, A. and Ketema, S. (2006) Strategic priorities for agricultural development in Eastern and Central Africa. Research Report 150, International Food Policy Research Institute, Washington, DC. Shannon, C. (1948) A mathematical theory of communication. Bell System Technology Journal, 27( 1948): 379—423. Siebert, Stefan; Döll, P. and Hoogeveen, J. (2001) Global map of irrigated areas version 2.0. Center for Environmental Systems Research, University of Kassel, Germany / Food and Agriculture Organization of the United Nations , Rome, Italy. Staal, S.J., Baltenweck, I., Waithaka, M.W., deWolff, T. and Njoroge, L. (2002) Location and uptake: integrated household and GIS analysis of technology adoption and land use, with application to smallholder dairy farms in Kenya. Agricultural Economics, 27, 295—315. Wood, S. and Chamberlin, J. (2003) Enhancing the role of spatial analysis in strategic impact assessment: improving data resolution for regional studies. Quarterly Journal of International Agriculture, 42(2): 167—187. Wood, S., Sebastian, K., Nachtergaele, F., Nielsen, D. and Dai, A. (1999) Spatial aspects of the design and targeting of agricultural development strategies, Environment and Production Technology Division Discussion Paper No. 44, International Food Policy Research Institute, Washington, DC. 159 You, L. and Wood, S. (2003) Spatial allocation of agricultural production using a cross-entropy approach. Environment and Production Technology Division Discussion Paper No. 126, International Food Policy Research Institute, Washington, DC. You, L. and Wood, S. (2006) An entropy approach to spatial disaggregation of agricultural production. Agricultural Systems, Vol. 90, Issues 1—3, 329—347. You, L.; Wood, S. and Wood-Sichra, U. (2007) Generating plausible crop distribution and performance maps for sub-Saharan Africa using a spatially disaggregated data fusion and optimization approach. EPTD Discussion paper 154, International Food Policy Research Institute, Washington, DC. Zellner, A. (1988) Optimal information processing and Bayes Theorem . American Statistician, 42, 278—284. Abstract</meta-value></custom-meta></custom-meta-wrap></article-meta></front><back><notes><p><list list-type="order"><list-item><p>1. Spatial data are sometimes referred to as geo-referenced data. Spatial data include some specific reference to location on the surface of the earth, expressed directly or indirectly through geographical coordinates, such as latitude and longitude. An important implied attribute of `spatial' data in economic studies is that of disaggregation to sub-national levels.</p></list-item><list-item><p>2. In this paper, `sub-national unit' refers to the first geopolitical division within a country, such as districts in Uganda, regions in Nigeria, or provinces in South Africa. Second level sub-national data are rarely available for SSA.</p></list-item><list-item><p>3. The model inputs and output described in this article all use a common grid resolution of 5 by 5 minutes of latitude and longitude. At the equator, this is equivalent to approximately 9.25 by 9.25 kilometers, or roughly 9000 hectares. All data used in our model are defined on the basis of the World Geodetic System 1984 Datum.</p></list-item><list-item><p>4. More details can be found under Global Agro-Ecological Zone 2000 at http://www.fao.org/ag/AGL/agll/gaez/ index.htm</p></list-item><list-item><p>5. A <italic>Wikipedia</italic> entry on cross entropy (http://en.wikepedia. org/wiki/Cross-entropy_method) may be helpful.</p></list-item><list-item><p>6. Shannon (1948) introduced information entropy to measure the uncertainty of the information in a given message, work which later gave rise to information theory. Information theory maintains that uncertainty is a statistical property of a message. Jaynes (1957) adopted the information entropy concept and proposed the maximum entropy principle in statistical inference: the least informative probability distribution can be found by maximizing the entropy of that distribution. In other words: in the absence of information to the contrary, all possible states of system are equally likely. Maximizing entropy is in fact a special case of minimizing cross-entropy with respect to a uniform distribution. The unique feature of the entropy approach is to overcome two empirical problems that hamper traditional econometrics: multi-collinearity and partially incomplete data (Golan, Judge and Miller, 1996). The idea is to remove irrelevant information at the beginning of a problem rather than taking pains to make dubious assumptions. Zellner (1988) described the advantage of the entropy approach as satisfying the `information conservation principle', by neither ignoring relevant input information nor injecting any false information.</p></list-item><list-item><p>7. Detailed model and equations could be found at You, Wood and Wood-Sichra (2007).</p></list-item><list-item><p>8. While we are not claiming that the reported agricultural statistics (either national or sub-national) are more accurate than, say, cropland indications derived from satellite imagery, we do establish the reported statistics as the basis for the ultimate crop maps. Thus, all the output grid cells corresponding to a given reporting unit (say, a province) will sum to that unit's reported total.</p></list-item><list-item><p>9. Except for some island countries such as Mayotte, Seychelles which have no or little agricultural production.</p></list-item><list-item><p>10. See Wood and Chamberlin, 2003 for fuller discussion of the value of spatially disaggregated agricultural data for rural development analysis and planning.</p></list-item></list></p></notes><ref-list><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Anselin, L.</surname></name> (<year>2002</year>) <article-title>Under the hood: issues in the specification andinterpretationofspatialregressionmodels</article-title> . <source>Agricultural Economics</source>, <volume>27</volume>, <fpage>247</fpage>—<lpage>267</lpage> .</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Bell, K.</surname></name> and <name name-style="western"><surname>Irwin, E.G.</surname></name> (<year>2002</year>) <article-title>Spatially explicit micro-level modeling of land use change at the rural-urban interface</article-title> . <source>Agricultural Economics</source> , <volume>27</volume>, <fpage>217</fpage>—<lpage>232</lpage> .</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Center for International Earth Science Information Network (CIESIN)</surname></name> , <name name-style="western"><surname>Columbia University</surname></name> ; <name name-style="western"><surname>International Food Policy Research Institute (IFPRI) and World Resources Institute (WRI).</surname></name> (<year>2000</year>) <source>Gridded Population of the World (GPW), Version 2</source>. <publisher-loc>Palisades, NY</publisher-loc>: <publisher-name>CIESIN, Columbia University</publisher-name> . Available at <uri xlink:type="simple">http://sedac.ciesin.columbia.edu/plue/gpw</uri>.</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Faostat.</surname></name> (<year>2005</year>) <uri xlink:type="simple">http://faostat.fao.org/default.aspx</uri> . Last accessed in October <year>2005</year>.</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Fischer, Gunther</surname></name> , <name name-style="western"><surname>Shah, M.</surname></name> , <name name-style="western"><surname>Velthuizen, H.</surname></name> and <name name-style="western"><surname>Nachtergaele, F.</surname></name> (<year>2001</year>) <source>Global agro-ecological assessment for agriculture in the 21st century</source>. <publisher-name>International Institute for Applied Systems Analysis</publisher-name>, <publisher-loc>Laxenburg</publisher-loc> , Austria.</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Food and Agriculture Organization of the United Nations (FAO).</surname></name> (<year>1981</year>) <article-title>Report of the Agro-Ecological Zones Project</article-title> . <source>World Soil Resources Report</source>, No <issue>48</issue>, Vol.<volume>1—4</volume>, <publisher-loc>Rome</publisher-loc>, <publisher-name>FAO</publisher-name> .</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Food and Agriculture Organization of the United Nations (FAO).</surname></name> (<year>2003</year>) <uri xlink:type="simple">http://www.fao.org/ag/AGL/agll/gaez/index.htm</uri>. Last accessed in July <year>2003</year>.</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Golan, Amos;</surname></name><name name-style="western"><surname>Judge, G.</surname></name> and <name name-style="western"><surname>Miller, D.</surname></name> (<year>1996</year>) <source>Maximum entropy econometrics: robust estimation with limited data</source>. <publisher-loc>New York</publisher-loc>: <publisher-name>John Wiley & Sons</publisher-name> .</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Jaynes, E.T.</surname></name> (<year>1957</year>) <article-title>Information theory and statistical methods I</article-title> . <source>Physics Review</source>, <volume>106</volume> (<year>1957</year>): <fpage>620</fpage>—<lpage>630</lpage> .</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Jaynes, E.T.</surname></name> (<year>1979</year>) <source>Where do we stand on maximum entropy</source>? In <source>The maximum entropy formalism</source>. <name name-style="western"><surname>R. D. Levine</surname></name> and <name name-style="western"><surname>M. Tribus</surname></name> (eds.) <publisher-name>MIT Press</publisher-name>, <publisher-loc>Cambridge, MA</publisher-loc> , p.<fpage>15</fpage>.</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Kherallah, M.</surname></name> , <name name-style="western"><surname>Delgado, C.</surname></name> , <name name-style="western"><surname>Gabre-Madhin, E.</surname></name> , <name name-style="western"><surname>Minot, N.</surname></name> and <name name-style="western"><surname>Johnson, M.</surname></name> (<year>2000</year>) <article-title>The road half traveled: agricultural market reform in Sub-Sahara Africa</article-title> . <source>Food Policy Report.</source><publisher-name>International Food Policy Research Institute</publisher-name>, <publisher-loc>Washington, DC</publisher-loc> , USA.</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Luijten, J.C.</surname></name> (<year>2002</year>) <article-title>A systematic method for generating land use patterns using stochastic rules and basic landscape characteristics: results from a Colombian hillside watershed. Agriculture Ecosystem &</article-title><source>Environment</source>, <volume>95</volume>, <fpage>427</fpage>—<lpage>441</lpage> .</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Nelson, G.C.</surname></name> (<year>2002</year>) <article-title>Introduction to the special issue on spatial analysis for agricultural economists</article-title> . <source>Agricultural Economics</source>, <volume>27</volume>, <fpage>197</fpage>—<lpage>200</lpage> .</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Omamo, S.W.</surname></name> , <name name-style="western"><surname>Diao, X.</surname></name> , <name name-style="western"><surname>Wood, S.</surname></name> , <name name-style="western"><surname>Chamberlin, J.</surname></name> , <name name-style="western"><surname>You, L.</surname></name> , <name name-style="western"><surname>Benin, S.</surname></name> , <name name-style="western"><surname>Wood-Sichra, U.</surname></name> , <name name-style="western"><surname>Tatwangire, A.</surname></name> and <name name-style="western"><surname>Ketema, S.</surname></name> (<year>2006</year>) <source>Strategic priorities for agricultural development in Eastern and Central Africa. Research Report 150</source>, <publisher-name>International Food Policy Research Institute</publisher-name>, <publisher-loc>Washington, DC.</publisher-loc></citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Shannon, C.</surname></name> (<year>1948</year>) <article-title>A mathematical theory of communication</article-title> . <source>Bell System Technology Journal</source>, <volume>27</volume>( <year>1948</year>): <fpage>379</fpage>—<lpage>423</lpage> .</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Siebert, Stefan</surname></name> ; <name name-style="western"><surname>Döll, P.</surname></name> and <name name-style="western"><surname>Hoogeveen, J.</surname></name> (<year>2001</year>) <source>Global map of irrigated areas version 2.0</source>. <publisher-name>Center for Environmental Systems Research, University of Kassel</publisher-name> , Germany / <publisher-name>Food and Agriculture Organization of the United Nations</publisher-name> , <publisher-loc>Rome</publisher-loc> , Italy.</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Staal, S.J.</surname></name> , <name name-style="western"><surname>Baltenweck, I.</surname></name> , <name name-style="western"><surname>Waithaka, M.W.</surname></name> , <name name-style="western"><surname>deWolff, T.</surname></name> and <name name-style="western"><surname>Njoroge, L.</surname></name> (<year>2002</year>) <article-title>Location and uptake: integrated household and GIS analysis of technology adoption and land use, with application to smallholder dairy farms in Kenya</article-title> . <source>Agricultural Economics</source>, <volume>27</volume>, <fpage>295</fpage>—<lpage>315</lpage> .</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Wood, S.</surname></name> and <name name-style="western"><surname>Chamberlin, J.</surname></name> (<year>2003</year>) <article-title>Enhancing the role of spatial analysis in strategic impact assessment: improving data resolution for regional studies</article-title> . <source>Quarterly Journal of International Agriculture</source>, <volume>42</volume>(<issue>2</issue>): <fpage>167</fpage>—<lpage>187</lpage> .</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>Wood, S.</surname></name> , <name name-style="western"><surname>Sebastian, K.</surname></name> , <name name-style="western"><surname>Nachtergaele, F.</surname></name> , <name name-style="western"><surname>Nielsen, D.</surname></name> and <name name-style="western"><surname>Dai, A.</surname></name> (<year>1999</year>) <source>Spatial aspects of the design and targeting of agricultural development strategies</source>, <source>Environment and Production Technology Division Discussion Paper No. 44</source>, <publisher-name>International Food Policy Research Institute</publisher-name>, <publisher-loc>Washington, DC.</publisher-loc></citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>You, L.</surname></name> and <name name-style="western"><surname>Wood, S.</surname></name> (<year>2003</year>) <source>Spatial allocation of agricultural production using a cross-entropy approach</source>. <source>Environment and Production Technology Division Discussion Paper No. 126</source>, <publisher-name>International Food Policy Research Institute</publisher-name>, <publisher-loc>Washington, DC.</publisher-loc></citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>You, L.</surname></name> and <name name-style="western"><surname>Wood, S.</surname></name> (<year>2006</year>) <article-title>An entropy approach to spatial disaggregation of agricultural production</article-title> . <source>Agricultural Systems</source>, Vol. <volume>90</volume>, Issues <issue>1—3</issue>, <fpage>329</fpage>—<lpage>347</lpage> .</citation></ref><ref><citation citation-type="book" xlink:type="simple"><name name-style="western"><surname>You, L.</surname></name> ; <name name-style="western"><surname>Wood, S.</surname></name> and <name name-style="western"><surname>Wood-Sichra, U.</surname></name> (<year>2007</year>) <source>Generating plausible crop distribution and performance maps for sub-Saharan Africa using a spatially disaggregated data fusion and optimization approach. EPTD Discussion paper 154</source>, <publisher-name>International Food Policy Research Institute</publisher-name>, <publisher-loc>Washington, DC</publisher-loc> .</citation></ref><ref><citation citation-type="journal" xlink:type="simple"><name name-style="western"><surname>Zellner, A.</surname></name> (<year>1988</year>) <article-title>Optimal information processing and Bayes Theorem</article-title> . <source>American Statistician</source>, <volume>42</volume>, <fpage>278</fpage>—<lpage>284</lpage> .</citation></ref></ref-list></back></article></istex:document></istex:metadataXml><mods version="3.6"><titleInfo lang="en"><title>Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</title></titleInfo><titleInfo type="alternative" lang="en" contentType="CDATA"><title>Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model</title></titleInfo><name type="personal"><namePart type="given">Liangzhi</namePart><namePart type="family">You</namePart><affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</affiliation><affiliation>International Food Policy Research Institute (IFPRI) in Washington, DC</affiliation><role><roleTerm type="text">author</roleTerm></role></name><name type="personal"><namePart type="given">Stanley</namePart><namePart type="family">Wood</namePart><affiliation>International Food Policy Research Institute</affiliation><affiliation>International Food Policy Research Institute</affiliation><role><roleTerm type="text">author</roleTerm></role></name><name type="personal"><namePart type="given">Ulrike</namePart><namePart type="family">Wood-Sichra</namePart><affiliation>International Food Policy Research Institute in Washington, DC</affiliation><affiliation>International Food Policy Research Institute in Washington, DC</affiliation><role><roleTerm type="text">author</roleTerm></role></name><name type="personal"><namePart type="given">Jordan</namePart><namePart type="family">Chamberlin</namePart><affiliation></affiliation><affiliation>E-mail: j.chamberlin@cgiar.org</affiliation><affiliation>International Food Policy Research Institute, j.chamberlin@cgiar.org</affiliation><role><roleTerm type="text">author</roleTerm></role></name><typeOfResource>text</typeOfResource><genre type="research-article" displayLabel="research-article"></genre><originInfo><publisher>Sage Publications</publisher><place><placeTerm type="text">Sage UK: London, England</placeTerm></place><dateIssued encoding="w3cdtf">2007-05</dateIssued><copyrightDate encoding="w3cdtf">2007</copyrightDate></originInfo><language><languageTerm type="code" authority="iso639-2b">eng</languageTerm><languageTerm type="code" authority="rfc3066">en</languageTerm></language><physicalDescription><internetMediaType>text/html</internetMediaType></physicalDescription><abstract lang="en">Agricultural production statistics are fundamental parameters for agriculture policy research. Information on acreage and yields of important crops is critical for understanding trends within what is the most important economic sector of many developing countries. Sub-national data — i.e. data organized by administrative units such as regions or districts — enable the analysis of patterns within countries that may highlight important policy issues, such as the need to allocate resources to underproductive areas. However, collecting sub-national data is difficult for developing countries with limited resources. Even with great effort, and often only on broad regional scales, enormous data gaps exist and are unlikely to be filled. As a result, information is often only available at national or very broad sub-national levels (such as provinces). Such geographically coarse data are unable to reflect important variations within countries and are insufficient for the spatial analysis of production patterns and trends. To fill these spatial data gaps we developed a model to disaggregate production data from coarser to finer spatial units. Using a cross-entropy approach, our spatial allocation model attempts to make plausible allocations of crop production from large reporting units such as a country or state, into smaller spatial units organized as cells of a regularly-spaced grid. In addition to more detailed information, the organization of production information in geographic grids allows for greater analytical possibilities through geographic information systems. The allocation model works on the basis of available evidence of mapped indicators of agricultural production, which include farming systems, land cover, crop biophysical suitability surfaces, commodity prices and local market access. This article describes the generation of crop distribution maps for Sub-Saharan Africa for the year 2000 using the spatial allocation model and discusses the importance of such maps for development analysis and planning.</abstract><subject><genre>keywords</genre><topic>cross entropy</topic><topic>spatial allocation</topic><topic>agricultural production</topic><topic>crop suitability</topic><topic>geographic information systems</topic><topic>Sub-Saharan Africa</topic></subject><relatedItem type="host"><titleInfo><title>Information Development</title></titleInfo><genre type="journal">journal</genre><identifier type="ISSN">0266-6669</identifier><identifier type="eISSN">1741-6469</identifier><identifier type="PublisherID">IDV</identifier><identifier type="PublisherID-hwp">spidv</identifier><part><date>2007</date><detail type="volume"><caption>vol.</caption><number>23</number></detail><detail type="issue"><caption>no.</caption><number>2-3</number></detail><extent unit="pages"><start>151</start><end>159</end></extent></part></relatedItem><identifier type="istex">7B56D53FD66612A3E586489E71BA0D2687E53E51</identifier><identifier type="DOI">10.1177/0266666907078670</identifier><identifier type="ArticleID">10.1177_0266666907078670</identifier><recordInfo><recordContentSource>SAGE</recordContentSource></recordInfo></mods></metadata><serie></serie></istex></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Agronomie/explor/SisAgriV1/Data/Istex/Corpus
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 001312 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Istex/Corpus/biblio.hfd -nk 001312 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Wicri/Agronomie
|area= SisAgriV1
|flux= Istex
|étape= Corpus
|type= RBID
|clé= ISTEX:7B56D53FD66612A3E586489E71BA0D2687E53E51
|texte= Generating Plausible Crop Distribution Maps for Sub-Saharan Africa Using a Spatial Allocation Model
}}
| This area was generated with Dilib version V0.6.28. |  |