The main caution to bear in mind when using SPI maps is that they are based on preliminary data from the Climate Prediction Center (CPC) and National Climatic Data Center (NCDC) rather than on final data. This is the same preliminary data used to computer the weekly Palmer Drought Index. It takes three or four months to assemble final, quality-controlled data. Using preliminary data for the SPI makes it possible to use it as part of a drought early-warning and monitoring system.
Changes from preliminary to final versions of data (actual values) are generally minimal and would rarely affect the values shown on an SPI map. However, preliminary data are based on measurements gathered from 450 to 550 stations nationwide each month. Final data includes measurements from more than twice as many reporting stations. Stations tend to be closer together in the eastern and central states, so there might be one or two stations per climate division. But in the west, some climate divisions may have no stations reporting in a particular month, or may lack a first- or second-order station altogether, so SPI values are interpolated from surrounding divisions.
This means that local anomalies are not as well reflected by preliminary data as they would be by final data. Colorado and some other western states supplement the preliminary data from the CPC with a network of stations throughout the state to produce SPI maps based on individual site data (producing contours) rather than on the National Oceanic and Atmospheric Administration’s (NOAA) climate divisions. Those SPI maps are part of the state’s drought early warning system. The National Drought Mitigation Center is working to compute a contour-based SPI map first for Nebraska, where the data is nearly available, and then for other regions in the United States.
Data is gathered in the first few days of each month on a dial-in system at CPC from first- and second-order reporting stations, mainly operated by NOAA, the Federal Aviation Administration, and military installations. The National Climatic Data Center (NCDC) averages departures from normal at each station to come up with the departure from normal for a climate division, and then uses normal values for each division to compute actual temperature and precipitation values. Then the Western Regional Climate Center downloads the data and computes the SPI for each climate division using the divisional precipitation values generated at NCDC. Finally, the data are transferred to the NDMC, where we map it and do additional analysis with a Geographic Information System. The NDMC is currently producing national SPI maps by climatic division. These maps are updated each month and are usually available and online by the 15th of the month.
The data set we use is the first of four updates that are applied to the climate division data base at the NCDC. It is also known as the preliminary operational data set. The second through the fourth updates occur anywhere from six weeks to three or four months after the original data month and include cooperative (COOP) station data. By the fourth update, all of the data have been through some sort of quality control.