CLIMATE VARIABILITY, AGRICULTURE AND REMOTE SENSING - Food Security, Food Prices and Climate Variability - page 63

0.9

0.8 Peak height

0.7

0.6

0.5

Start of season

End of season

0.4

0.3

Length of the season

0.2

0.1

0

0 50 100 150 200 250 300 350 400

Day of the year

FIGURE 3.5 Phenology metrics derived from models using NDVI temporal curves, including the
start of season, end of season, length of season, peak height, peak position in days of
the year and cumulative NDVI.

strategies. These changes meant that the methods typically used by food security analysts, that
use yield estimate models to assess the relative impact of weather on food production, were less
useful. For example, the yield model used by FEWS NET is the water requirement satisfaction
index or WRSI model (Senay and Verdin, 2002, 2003). It uses rainfall as an input and expresses
yield as a percent of normal. If the “normal” or average conditions no longer represent actual
production because of a wholesale change in farming practices, then these models are not very
useful. Thus the approach taken by Funk and Budde in the 2009 paper was essential as well as
effective in providing a new baseline for estimating production.
Phenology and NDVI assessments allow production of national assessments of production
deficits halfway through the growing season, which will assist humanitarian organizations with
their planning during a crisis. Zimbabwe was particularly ill-equipped to deal with weather-
related production declines in 2008 due to its lack of foreign exchange and hostile rhetoric
towards the United States and Europe, restricting the ability of aid organizations to assist.
Delayed planting date, lack of fertilizer and modern crop strains, as well as inadequate rainfall
were the foundations for a significant food security crisis in Zimbabwe in 2008. The analysis that
was conducted for FEWS NET using phenology curves from remote sensing data provided a
very highly accurate, early and spatially explicit estimate of production when there was consider-
able uncertainty as to how much aid Zimbabwe would require (Funk and Budde, 2009b).

Socio-economic factors and the impact of climate variability on
agriculture

How do socio-economic factors influence how harvests are affected by changes in temper-
ature and precipitation? Although rainfed agriculture is always sensitive to large changes in
growing conditions, some regions are more sensitive than others. Recent research has found