National Academies Press: OpenBook

People and Pixels: Linking Remote Sensing and Social Science (1998)

Chapter: 9 Social Science and Remote Sensing in Famine Early Warning

« Previous: 8 Extraction and Modeling of Urban Attributes Using Remote Sensing Technology
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 189
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 190
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 191
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 192
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 193
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 194
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 195
Suggested Citation:"9 Social Science and Remote Sensing in Famine Early Warning." National Research Council. 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, DC: The National Academies Press. doi: 10.17226/5963.
×
Page 196

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

9 Social Science and Remote Sensing in Famine Early Warning Charies F. Hutchinson The specter of famine has haunted humans throughout history. Yet despite the havoc it has wreaked on various populations, the causes of famine are poorly understood. Famine has been viewed as the will of God, as an accident of nature, as an inexorable consequence of economics and, most recently, as a social pro- cess. Because it strikes a fundamental chord that links all of humanity, famine, or the threat of famine, is an important issue that occupies a more prominent place in popular thinking than issues such as high rates of malnutrition, morbidity, and mortality in the low-income countries of the world. Thus, it is generally accepted that there is a need to better understand the process of famine so it may be averted and ultimately eliminated. Formal efforts to monitor food security and predict famine began more than a century ago in India, where colonial administrators sought evidence of impend- ing food shortages in rural areas to avert disaster. While the resulting Famine Codes have persisted there, it was not until the great droughts of the late 1960s and early 1970s in the Horn and Sahel regions of Africa that the international community paid attention to the need to develop a more comprehensive system for routine monitoring of food security (Arnold, 1988~. Between 1975 and the present, several systems have evolved for monitoring food conditions, primarily in Africa. An increasingly sophisticated array of tools has been applied to the problem, including a number of satellite remote sensing techniques (Hutchinson, 1991~. However, the utility of these new tools has been hampered by inadequacies in (1) general understanding of the famine process, and (2) specific understanding of the interaction between restrictions in access to food and reactions to those restrictions among various groups. 189

90 SOCIAL SCIENCE AND REMOTE SENSING IN FAMINE EARLY WARNING The objectives of this chapter are to review different models of famine and how they have been adopted in various famine early warning systems, to consider the evolving role of remote sensing in famine early warning, and to explore how social science and information technologies (i.e., remote sensing and geographic information systems) appear to be converging in the famine early warning arena as a potential model for other integrated assessments. PRACTICAL IMPLEMENTATION OF FAMINE MODELS The major systems for monitoring food security or the likelihood of famine- operate at continental and global scales. The monitoring activities described here rely on secondary aggregate data (data routinely gathered and reported for admin- istrative districts by national governments) and on observations made by satel- lites. Because of the expense involved, primary field data are gathered only after a country or region within a country has been identified as a potential problem. The Food and Agriculture Organization (FAO) of the United Nations launched the Global Information and Early Warning System (GIEWS) in 1975. At that time, famine was defined in Malthusian terms as a situation in which demand for food exceeds supply (i.e., failure of supply). The tool developed to describe food status was a national "food balance sheet" in which food demand as a function of population size was weighed against agricultural production for the current year, plus scheduled imports and food stocks carried over from previous years. Although fairly crude, the food balance sheet came into general use because it could be developed relatively early in the growing season and yielded a product that was easily understood and acted upon (Global Information and Early Warning System, 1997~. The return of catastrophic droughts in the Horn and Sahel regions of Africa in 1984 again drew international attention to the problem of famine. At the Bonn Summit of 1985, donor countries agreed to renew their collective efforts to detect and avert future famine, particularly in Africa. This renewed interest spurred developments in famine early warning, and efforts were made to improve the food balance sheet approach pioneered by GIEWS. Almost all suggested refine- ments to the approach involved improving the supply side of the balance sheet, and national estimates of crop production became the obvious target. Two gen- eral approaches to improving crop production estimates emerged. One was based on models of crop yield driven by direct observations of precipitation. Although this approach was promising, its adoption was hindered by a lack of reliable meteorological data and an insensitivity to other factors that might affect total production (e.g., availability of inputs, estimates of planted area). The other, more direct approach was based on observing the progress of the growing season using remotely sensed data (Hutchinson, 1991~. Beginning in 1979, great strides were made in regional monitoring based on the use of satellite data, especially those provided by the National Oceanic and

CHARLES F. HUTCHINSON 191 Atmospheric Administration (NOAA) series of satellites carrying the Advanced Very High Resolution Radiometer (AVHRR) instrument. With these data it was possible to monitor conditions routinely for all of Africa, summarized over a standardized 10-day period (dekad), for pixels with an area of about 16 km2. While this coarse scale does not allow statements to be made about specific fields or communities, it does provide an overview of how the growing season is pro- gressing for an area that may contain several communities or for an administra- tive district. Similarly, it is difficult at such a scale to derive quantitative esti- mates of crop yields for specific areas. However, the development of an archive of observations since 1981 has made it possible to compare a given period of the current year with the same period during the past year or with mean conditions over the period of record, and thus to develop a reasonably reliable estimate of the quality of the growing season and anticipated crop yield. Thus, it is possible to assert with some confidence whether the cropping season in an area will be better or worse compared with last year or the average. Two programs to incorporate satellite data in famine early warning were launched in 1985: FAO developed the Africa Real Time Environmental Moni- toring Information System (ARTEMIS) to provide data to GIEWS and other regional monitoring programs (e.g., locust control operations) for all of Africa (Hielkema et al., 1986), while the U.S. Agency for International Development (USAID) developed the Famine Early Warning System (FEWS), which operated initially in the Sahel and Horn of Africa (Hutchinson, l991~. Both programs use and share AVHRR data processed by the National Aeronautics and Space Ad- ministration (NASA) to describe vegetation conditions, and both have added rainfall estimates based on cloud observations derived from the European Meteosat satellite (Snijders, 1991~. Despite improved estimates of food production from satellite data, it became obvious that famine was more complex than a simple failure of food supply. Studies of famine during the 1970s revealed that food was available during these events, yet people still starved (Garcia, 1981~. It was proposed that in many, if not most, emergency situations, food may be available, but the mechanisms of exchange (entitlements) by which people have traditionally gained access to food cease to function (See, 1981~. Thus, rather than a failure of supply, famine is caused by a failure of effective demand. The need to gather, monitor, and analyze data on food access presented new challenges. FEWS initially adopted a "convergence-of-evidence" approach used routinely in air photo and satellite image interpretation (Estes et al., 1983~. In addition to information derived from images, this approach considers a wide array of data types, with an interpretation being made that is supported by all the data. In the FEWS approach, a minimum of three sets of information, or classes of "indicators," was considered. The first set included those factors that might suggest food supply (e.g., observed precipitation; early season yield estimates; reports of the incidence of pests, such as locusts; and estimates of food stocks in

92 SOCL4L SCIENCE AND REMOTE SENSING IN FAMINE EARLY WARNING granaries carried over from the preceding year). The second set comprised those factors that might suggest food access (e.g., cereal prices; small animal prices; terms of trade between cereal and animals; labor prices; and anecdotal informa- tion, such as reports by travelers about conditions in outlying areas). The third set included those factors that might indicate levels of development (e.g., access to roads, water, and health and school facilities). These sets of data were then interpreted together. For example, if the growing season was below average for a district, but stocks were high, the area was accessible by road, and terms of trade held steady, convergence of evidence suggested that an emergency was unlikely. Conversely, if an indifferent year followed 2 years of poor harvest in an isolated area and terms of trade began to deteriorate quickly, convergence of evidence suggested that the poorest households would have restricted access to food, and the likelihood of an emergency was high. Adoption of the convergence-of-evidence approach was an advance in early warning because it increased confidence in the assessment. It also offered a number of other benefits because it led to the development of large, dynamic, general-purpose databases that have application beyond early warning. For ex- ample, the numbers of health care clinics or schools in a district suggested levels of development that might also be used to infer ease of access to food. Similarly, distance to primary roads would indicate the level of development and thus access to markets or alternative sources of income. While indicative of vulner- ability to food security emergencies, these same data could be used as guidelines for long-term development (e.g., to determine where clinics, schools, or roads were most needed). While the convergence-of-evidence approach offered a relatively sensitive and reliable indication of current conditions, it still provided little information on the differential impacts an emergency might have on various groups or types of households (e.g., pastoralists, subsistence farmers, female-headed households). Gaining an understanding of the economic and social contexts in which an emer- gency, such as a drought, might play out at the household level became and remains a significant challenge. Efforts continued to focus on gathering and analyzing secondary aggregate data, but there was a gradual shift from the simple convergence-of-evidence approach to an approach that attempted to interpret data with regard to how they might reflect household response to current conditions. Watts (1983) and later Corbett (1988) offered a simple conceptual model in which the way households react to emergencies or perceived threats to their economic condition is deter- mined by considering how they commit household resources. A modification of this model was adopted by FEWS; see Figure 9-1 (Famine Early Warning Sys- tem, 1997~. In this model, famine is viewed as a process that unfolds over time. Monitoring data can be placed within this framework so that conditions at the household level can be inferred from aggregate data. For example, terms of trade can be expected to deteriorate fairly early in an emergency and continue to

CHARLES F. HUTCHINSON DONOR RESPONSES - Developed ~ .-"ing~on~ _ t Relief ~ HOUSEHOLD VElLPlE~B~TY HOUSEHOLD STRATEGIES 1 Adaption Diet change' borrowings scasonat labor m~gra~on D'wstmcnt 1 Liquid its 1 Productive wets al . I., of _ c to ·e ~ ~-~1 1 \ l rune FIGURE 9-1 Household responses to the threat of food security emergencies. 193 l Crop ~ LiveStoCIc Adjustments ~Dict change 1 Leonine food usp ~` Grain loan from Icin 1 1 Labor sales (migration) ~ Small animal sales ~ \ \ Cash~cereal man From merchants A \<Produciive asset sales \ Farmland pledging Farmland sale ~ Ou~niglation ~\

94 SOCIAL SCIENCE AND REMOTE SENSING IN FAMINE EARLY WARNING decline as grain prices rise because of demand and as the relative prices of small animals (e.g., goats) fall because more are offered for sale. Further, it can be inferred that an emergency is quite serious when resources that are critical to household livelihoods (e.g., draft animals, farm implements) begin to appear in markets, and that a crisis has arrived when large-scale migrations to urban areas and/or refugee centers begin to take place. OPPORTUNITIES FOR INTEGRATION OF SOCIAL SCIENCE Continued work on the basic nature of vulnerability to famine (Watts and Bohle, 1993) and how this vulnerability is manifested among various groups (Seaman, 1997) has opened opportunities for further integration of remote sens- ing and social science in famine early warning. Certainly there have been signifi- cant advances in the power with which aggregate data (e.g., district-level produc- tion data) can be manipulated for example, using geographic information systems and in the ways these data can be combined with other spatially refer- enced observations (e.g., subdistrict-level land use derived from remote sensing data) to yield more geographically specific results. Nevertheless, there are very real limits on inference that cannot be overcome without a better understanding of the economic, social, and political differences among groups and households, and how these differences might vary in space and time. Few of these conditions can be revealed through remote sensing. In recognition of these limitations, FEWS has added some caveats to the interpretation of Figure 9-1 (Famine Early Warning System, 1997:12~: The ability of households to withstand emergencies is conditioned by (a) the depth, diversity and quality of their resource base, (b) the breadth of their in- come portfolio, and (c) their relationship to economic, social and political hier- archies. Consequently, in parallel with the efforts of FEWS to refine the use of quantitative aggregate data for inferring household responses to and the impacts of emergencies (a top-down approach), others in the early warning community have sought to develop a more basic qualitative understanding of household economies among various livelihood systems and wealth groups that might be extended over an area (a bottom-up approach). The Save the Children Fund (SCF) has developed a "food economy" approach based on structured interviews with key informants within "food economy zones" (Seaman, 1997~. The result- ing data are entered into a computerized model, which can then be subjected to various perturbations (e.g., simulation of food production losses through drought). In this way, an estimate of the impact on "household food income" is projected. This estimate is expressed differentially by wealth type of household, and the net food shortfall is mapped out in relation to the food economy zone. Thus where

CHARLES F. HUTCHINSON 195 FEWS, by necessity, applies a generic model of household behavior in all areas and to all groups, SCF develops a set of models that is area- and group-specific. The bottom-up approach based on specific household models also has its limitations, however. Models developed through field survey are essentially static. If significant changes occur within a region (e.g., if a new road is built or a mine is opened), the models must be revised, again through resource-intensive field survey. Also, the system is usually geographically restricted in its applica- tion: it tends to be used in designated areas where relief operations are planned, with less attention being paid to areas outside these zones, making it difficult to extend the findings over larger (e.g., multiethinic, multinational) areas. Finally, the numbers yielded by this approach are often viewed as "soft" in that they incorporate a good deal of qualitative or nonformal data derived from interviews with key informants, rather than a statistically rigorous sample. As a result, the top-down and bottom-up approaches have been judged to be complementary rather than competitive (Global Information and Early Warning System, 1997~. The maintenance of dynamic national databases makes it pos- sible to make statements about an entire country, and offers benefits that extend beyond famine early warning. Moreover, the limitations of a single, universal model of household response resulting from the restricted use of aggregate data can be overcome by developing region- and group-specific models through field survey. With a more specific understanding of household behaviors, it will be possible to offer differential interpretations of the effects of an emergency on different households and different areas using the same monitoring data. At present, discussion of the involvement of social science in the food secu- rity community is restricted primarily to what the fields of anthropology and geography might contribute to an improved understanding of household behav- ior, largely through the use of rapid qualitative field survey methods combined with the use of geographic information systems to provide a link to routinely reported aggregate data. However, the potential contribution of social science extends beyond this level to issues of theory and the use of more quantitative techniques to achieve better understanding and prediction of human behavior. The need for collaboration with the social sciences has been acknowledged and will occur, at least within this initial context (Global Information and Early Warning System, 1997~. Once the fundamentally social nature of the famine process is acknowledged, expansion of this collaboration will be recognized as the most fruitful frontier for early warning research. The greatest challenges to such increased collaboration may well lie in bridging the cultural gaps that exist between academic disciplines. The famine early warning community has recog- nized these challenges and the need to resolve them.

96 SOCL4L SCIENCE AND REMOTE SENSING IN FAMINE EARLY WARNING REFERENCES Arnold, D. 1988 Famine: Social Crisis and Historical Change. Oxford, England: Basil Blackwell. Corbett, J. 1988 Famine and household coping strategies. World Development 16(9):1099-1112. Estes, J.E., E.J. Hajic, and L.R. Tinney 1983 Fundamentals of image analysis: Analysis of visible and thermal infrared data. Pp. 987- 1124 in Manual of Remote Sensing, Second Edition, R.N. Colwell, ed. Washington, D.C.: American Society of Photogrammetry. Famine Early Warning System (FEWS) 1997 Vulnerability analysis and FEWS. Annex 3/2 in Global Information and Early Warning System (GIEWS). Summary Report: Second Informal Meeting on Methodology for Vul- nerability Assessment. ES:GCP/INT535/EEC. Rome, Italy: Food and Agriculture Orga- nization of the United Nations. Garcia, R.V. 1981 Drought and Man: The 1972 Case History, Vol. 1. Nature Pleads Not Guilty. Oxford, England: Pergamon Press. Global Information and Early Warning System (GIEWS) 1997 Summary Report: Second Informal Meeting on Methodology for Vulnerability Assess- ment. ES:GCP/[NT535/EEC. Rome, Italy: Food and Agriculture Organization of the United Nations. Hielkema, J.U., J.A. Howard, C.J. Tucker, and H.A. Van Ingen Shenau 1986 The FAO/NASA/NLR ARTEMIS system. Pp. 147-160 in Proceedings: Twentieth Inter- national Symposium on Remote Sensing of Environment. Ann Arbor: Environmental Research Institute of Michigan. Hutchinson, C.F. 1991 Uses of satellite data for famine early warning in sub-Saharan Africa. International Journal of Remote Sensing 12(6):1405-1421. Seaman, J. 1997 The food economy approach to vulnerability assessment and the RiskMap computer pro- gram. Annex 3/1 in Global Information and Early Warning System (GIEWS). Summary Report: Second Informal Meeting on Methodology for Vulnerability Assessment. ES:GCP/[NT535/EEC. Rome, Italy: Food and Agriculture Organization of the United Nations. Sen, A.K. 1981 Poverty and Famine: An Essay on Entitlement Deprivation. Oxford, England: Clarendon Press. Snijders, F.L. 1991 Rainfall monitoring based on Meteosat data a comparison of techniques applied to the western Sahel. International Journal of Remote Sensing l 2: 1331 - 1347. Watts, M.J. 1983 Silent Violence: Food, Famine and Peasantry in Northern Nigeria. Berkeley: University of California Press. Watts, M.J., and H.G. Bohle 1993 The space of vulnerability: The causal structure of hunger and famine. Progress in Human Geography 17(1):43-67.

Next: 10 Health Applications of Remote Sensing and Climate Modeling »
People and Pixels: Linking Remote Sensing and Social Science Get This Book
×
Buy Paperback | $44.00 Buy Ebook | $35.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Space-based sensors are giving us an ever-closer and more comprehensive look at the earth's surface; they also have the potential to tell us about human activity. This volume examines the possibilities for using remote sensing technology to improve understanding of social processes and human-environment interactions. Examples include deforestation and regrowth in Brazil, population-environment interactions in Thailand, ancient and modern rural development in Guatemala, and urbanization in the United States, as well as early warnings of famine and disease outbreaks. The book also provides information on current sources of remotely sensed data and metadata and discusses what is involved in establishing effective collaborative efforts between scientists working with remote sensing technology and those working on social and environmental issues.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!