562 
qualitatively compared to the historic occurrences 
of crop failure, food shortages, drought, flooding 
and other anomalous weather/non-weather events as 
determined from reports newspapers, and computerized 
data bases. Episodic data are also used to 
"calibrate" the index by establishing a critical 
threshold which is associated with crop failure 
and/or drought-related food shortages. The 
strengths and limitations of all available indices 
and models are collectively considered in the 
assessment process. 
Documented food shortages determined by in-country 
field inspections reported by the UN/FAO and 
American Embassies during 1979-1983 confirmed that 
AISC assessments consistently provide a 3-6 month 
alert on potential food shortage situations. 
Typically, the initial AISC assessment on 
drought/food shortage conditions was issued 30-60 
days before the beginning of the crop harvest. 
Actual indications of food shortage conditions were 
not evident until 2-4 months after harvest. 
3 UPGRADED CLIMATIC IMPACT ASSESSMENTS FOR AFRICA 
The NOAA/NESDIS Early Warning Program for Africa has 
been upgraded as the result of recent advances in 
the application of NOAA polar orbiting satellite 
data for agricultural assessment. Color-coded 
images and vegetation/biomass indexes are derived 
from the NOAA AVHRR sensor (Advanced Very High 
Resolution Radiometer) using daily four kilometer 
GAC data (i.e., Global Area Coverage). These 
satellite products have substantially upgraded 
AISC's ability to detect drought, determine its 
regional extent between sparse rainfall reporting 
stations, and assess relative vegetation/biomass 
conditions. Improved assessments of weather impacts 
on crops and rangelands are based on the integrated 
use of agroclimatic/rainfall index models and NOAA 
AVHRR satellite models. The integration of 
agroclimatic and satellite derived assessment 
information represents a classic opportunity for 
using a Geographic Information Systems (GIS) 
approach. In fact, AISC staff in Columbia, MO and 
Washington, D.C. are using a "light table" system to 
overlay and analyze various data for operational 
assessment of weather impacts on African 
agriculture. 
The techniques presently used for processing 
information in interpretative visual analysis are 
time consuming endeavors which need to be automated, 
i.e., computerized. Drawbacks of the manual 
technique are unavoidable due to the sheer volume of 
spatial, temporal and thematic (special topic) data, 
along with the scale at which they are obtained. 
There is an urgent need to embrace a new generation 
of spatial analysis for climatic impact assessment, 
namely automated Geographical Information Systems 
(GIS). 
To gain a better understanding of our diversified 
planet, scientists must gather, store, and analyze 
tremendous volumes of data on a wide range of 
subjects. Information needed in spatial analysis 
can range from bedrock structure, soils, weather 
observations, natural vegetation to population 
distribution, land use, political, economic and 
social structures. Understanding our world in terms 
of the physical and cultural landscape entails 
identifying and analyzing numerous relationships 
resulting from human and environmental interactions. 
The range, scope, and magnitude of information 
necessary for analyzing this world of the physical 
environment, socio-economic systems, and historical 
ties can become overwhelming. 
In recent years government agencies have 
recognized the potential of Geographic Information 
Systems (GIS) as a technology suited for rapid data 
manipulation. GIS system, therefore, are viewed as 
an integral component in spatial problem analysis 
and assessment. 
4 DEFINING GIS 
A Geographic Information System is the overlaying of 
spatial data (i.e., population density, land cover 
types, land ownership, transportation network) for 
the same geographic space at a uniform scale. In a 
more sophisticated expression, a GIS is geocoded 
information in a spatial and/or tabular format with 
manipulation capabilities of georeferenced 
relationships. 
Automated GIS are the results of methodological 
changes (i.e., digital data display in image format; 
superimposing of two georeferenced images) brought 
about by advances in technology. James and Martin 
(1981, p. 405) observe that: 
. . . changes in the technology of observations and 
analysis not only provide spatial scientists 
(geographers) with more information than has even 
been available before, but also provide a means of 
storage and recall, and a way to carry out complex 
analysis. 
The advent of automated or computerized data 
processing techniques in the 1950's and 1960's 
revolutionized the capabilities of creating and 
handling all types of information in problem solving 
procedures. With the aid of computers, vast amounts 
of data can be analyzed quickly and efficiently. 
Computers provided new means to analyzing complex 
problems through identifying relationships between 
variables never before thought possible. It was in 
the environment of the "quantitative revolution" 
that geographic data processing systems and GIS 
evolved. 
An automated GIS is a system that has as its 
primary source of input a base composed of 
geographic referenced data coordinates, and the 
majority of processing is done digitally with a 
computer. This digital GIS can be viewed as a 
multilayered vertical structure of related 
georeferenced data sets designed for use in solving 
spatial problems in a computer environment (fig. 1). 
Further, GIS are capable of analyzing large volumes 
of data acquired from a variety of sources (Marble, 
Peuquet and Calkins, 1984).