Food and Agriculture Organization and Agricultural Droughts - Monitoring and Predicting Agricultural Drought: A Global Study - page 414

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46

U se of Satellite Data
Two main satellite-based variables used in the FAO crop forecasting ap-
proach are NDVI and cold cloud duration (CCD). In FAO food security
programs, rainfall is often estimated from CCD. Low values of NDVI cor-
respond to sparse or no vegetation (ochre-brown-green), and high values
indicate dense vegetation (red-pink-purple). The CCD is an indicator based
on high frequency (hourly) thermal infrared observations from geostation-
ary meteorological satellites of the METEOSAT type. It is a measure of
the duration, in hours, in which clouds become so cold (below - 40° C)
that the likelihood of their producing rain is very high. The CCD has been
effectively used for agricultural drought monitoring because it serves as
a proxy for rainfall. This gives a picture of whether the current season is
better or worse than any previous season. However, its greatest potential is
to provide and map the estimates of rainfall, which ultimately affects agri-
cultural production in many parts of the world where agriculture is heavily
dependent on rainfall, as is the case in the sub-Saharan Africa. The Early
Warning System of the SADC has been using El Ni no/Southern Oscillation
Index values (SOI; chapter 3) to predict the likely outcome of the coming
rainy season (FAO, 1996).

[414

Line
——
-0.1
——
Norm
PgEn

Q uantitative Application of Satellite Imagery

Satellite imagery has been used for agricultural drought monitoring and
other activities for many years. However, less progress has been made in
the quantitative use of remote-sensing imagery. For example, the NDVI-
based analysis conducted by FAO, at continental and regional scales, has
not changed much since 1988. Although data registration and calibration
have become much better, the main type of analysis still is a qualitative
assessment of the current vegetation situation during the growing season,
as compared to previous years or the average, by comparing trends in
NDVI.
The use of METEOSAT data for rainfall estimates has also only been
partially successful. Although quantitative estimates of rainfall derived
from METEOSAT are becoming more refined, the most popular
METEOSAT-derived product for early warning and drought monitoring
remains the CCD images. These images were originally intended to be an
intermediate product only and not suitable for distribution. More recently,
METEOSAT data have been merged with data from other sources to im-
prove the quality of the estimates. At the national level, this is often done
by interpolating between a large number of observations from meteoro-
logical stations using the CCD images as a weighted surface, guiding the
interpolation patterns and the ground observations for quantification. At
a continental scale, only ground-observed rainfall data through the Global
Telecommunication System (GTS) is readily available in real time, which
often is of uncertain quality in many food-insecure countries. The sparse

[414