10.4
Seasonal Forecasts for Use in Weather Risk Management
Ants Leetmaa, NOAA/NWS/NCEP/CPC, Camp Springs, MD
The Climate Prediction Center (CPC) routinely produces seasonal forecasts for temperature and rainfall out to a year in advance. These are for shifts in probability anomalies for the categories of "below or above normal" and "near normal" for broad regions of the country. Modest levels of skill have been achieved especially for seasons when strong El Nino and La Nina impacts were expected. The shifts are relative to a climatology for the 1961-1990 period.
The past few years have seen rapid growth in the use of climate information, data and forecasts, in risk management practices, i.e weather derivatives, to lessen the economic impacts of adverse weather/climate fluctuations primarily in the utility industry. It has become clear that refinements are needed to make the CPC forecasts more relevant to the weather risk management community. Most users have a "memory" of their vulnerability to weather/climate variability that goes back 10-15 years. With the strong warming trends experienced over the past 30 years, the use of an "old" climatology compounds the difficulty of making accurate and useful forecasts. Weather derivative contracts are based on information observed usually at urban sites. The variability observed locally is not totally representative of the larger scale conditions which are being forecast and can be influenced by urban effects.
The CPC is working to implement improvements to its forecasts and to provide tools to the private sector so that some of these limitations can be overcome. A "probability of exceedence" forecast product has been produced so that users have access to the whole forecast probability distribution. Along with the graphical presentation, the forecast and climatological data can be downloaded for quantitative computations. Regression equations can also be downloaded which related the broad scale forecasts to most U.S. cities. For the coming winter forecast work is underway to provide forecast information for both of the low frequency variability, i.e. the trends, and the stronger season to season changes resulting from climate variability.
Session 10, Weather Risk and Derivatives (Parallel with Sessions 11 & 12)
Wednesday, 17 January 2001, 1:30 PM-4:30 PM
Previous paper Next paper