Because food consumption is difficult to measure, estimates of access have been based on
the difference between the price of food and the income of consumers. Rapid changes in the
price of food can be used to understand changing access to food due to market factors since
income rarely keeps pace. Databases of the price of food that is actually consumed by the
poorest people were developed after the 2008 food crisis, and early warning organizations
have begun to report the current price, in US dollars and local currency, with a calculation of
the percent change from the previous month, from three months ago and from the same time
in the previous year (Eilerts, 2013). Thus food prices are becoming far more central to how
food security analysis is conducted, and their impact on local food security far more widely
Climate variability and environmental observations
Climate variability describes the recent weather and trends in temperature and precipitation,
including the fluctuations associated with large-scale natural climate phenomenon such as El
Nino Southern Oscillation or ENSO. Extreme droughts and wet periods, due to the inter-
action of large-scale wind and sea surface temperature changes, are also referred to as climate
variability. Decadal and seasonal shifts in wind patterns and sea surface temperatures in the
Atlantic cause changes in hurricane frequency, for example. The dust bowl drought in the
1930s in the United States, usually attributed to climate variability, is another example of a
persistent and extreme climate event caused by larger decadal climate processes.
The interaction between climate variability and climate change is uncertain, since the
changes in mean climate and mean variability can occur simultaneously. For example, an
increase in the mean temperatures will give more extremes, but may not result in higher vari-
ability. Climate change may result in higher variability as well as a changed mean ( Plate 2 ).
Both of these changes will result in more extreme events that will have negative impact on
Extreme changes in temperature can be documented, but these extremes cannot easily be
connected to a change in the mean temperature, the mean variance or both. Uncertainties in
the rate of change of the mean also confound interpretation of changes in the variance, since
all variance statistics are dependent on the basis of estimate, which is the mean (Meehl, 2007).
Precipitation is even more complicated than temperature because it is often not normally
distributed, and thus statistical descriptions of the variability of rainfall are more complex, par-
ticularly over shorter time periods of five to ten years. The impact of environmental variabil-
ity and change on society is hard to estimate since concurrent and dramatic demographic, land
use and economic transformations have taken place over the same period.
The analysis in this topic uses to the extent possible the observational data record instead
of a modeling framework to understand the impact of climate variability on food security.
Unlike global or regional climate models, observations allow conclusions that are less suscep-
tible to model assumptions and model error. Observations are at a wide variety of spatial and
temporal resolutions, and enable the use of data appropriate to the analysis needs. Agricultural
regions have ine-scale gradients in temperature and precipitation that can have a significant
impact on growing conditions. Thus adequate spatial resolution is important for understand-
ing the impact of growing conditions on agricultural production.
Remote sensing observations can assess the impact of weather on agriculture and other
vegetation on the land surface. Repeat observations over decades can quantify the long-term