Drought Forecasting

Decision makers must have accurate drought monitoring information to respond successfully during drought events. Accurate drought forecast information and tools about future conditions are equally important. The science of drought forecasting, however, is in its infancy. To forecast drought, it is important to know something about the causes of drought. Drought is usually established by persisting high pressure that results in dryness because of subsidence of air, more sunshine and evaporation, and the deflection of precipitation-bearing storms. This is usually part of a persistent large-scale disruption in the global circulation pattern. Scientists are looking for local or distant influences that might create such atmospheric blocking patterns.

In March 2000, NOAA's CPC launched a new drought forecast tool for the United States called the Seasonal Drought Outlook (SDO) (http://www.cpc.ncep.noaa.gov/prod-ucts/expert_assessment/seasonal_drought.html). This tool is issued monthly at the same time as the traditional long-range seasonal temperature and precipitation forecasts. The SDO attempts to anticipate the pattern and trends for drought conditions across the country 3 months in advance. The development of the SDO incorporates a mix of tools, including statistical techniques based on historical natural and "constructed" analogues, historical drought index probabilities based on the time of the year, and various dynamical and statistical precipitation and temperature outlooks spanning various time periods. The SDO forecasts have met with mixed success so far, a major stumbling block being the difficulty of forecasting rainfall, or lack thereof, during the summer.

Given the sensitivity of the atmosphere to small-scale upper-level disturbances and the difficulty of forecasting such factors more than several days in advance, as well as the inherent "noisiness" of the rainfall pattern during summer, drought forecasts will be a challenge for a long time. That's the bad news. The good news is that continuing research on the mechanisms responsible for large-scale precipitation and drought patterns should lead to slow but useful improvements in seasonal forecasts and longer term forecasts. The forecasts will never attain the accuracy currently expected for short-term temperature and rainfall forecasts, but they should be better than flipping a coin ... maybe a lot better.

0 0

Post a comment