|
|
HOME >
Monitoring and Data >
U.S. Climate Data >
Precipitation & Temperature >
U.S. Temperature and Precipitation Trends >
Descriptions and Limitations
|
| |
|
| |
|
Descriptions and Limitations
|
| |
| These maps and bar graphs depict linear rates of change of seasonal and annual average temperatures and total precipitation since the mid-1970's (a) for each 3-month period (12 U.S. maps), (b) for the year as a whole (1 U.S. map), and (c) for the nation as a unit (bar graphs). Based on independent research by Livezey and Smith (1999), and supported by many other studies since, the Climate Prediction Center (CPC) is confident that the gross place-to-place and season-to-season variations in these trends are meaningful and statistically significant, and are related to global-scale trends in sea- and land-surface temperatures. Moreover, the attributes of these trends can be extrapolated into the near-future (the next year or two) to determine the approximate background climatology at that time. Further it can be shown (Livezey et al., 2007, in review) that the trend estimates are likely to be more representative than other estimates using common alternative methods, like linear trend fits and, for medium to large trends, OCN. However, these trends should not be used for longer time-period extrapolations because they cannot yet be confidently ascribed to specific physical processes and are probably only approximately linear. The possible causes of these observed trends are continually under investigation. An earlier version of these trends have been taken into account in both CPC's summary of U.S. La Niña Seasonal Temperature and Precipitation Impacts for 1999-2000, and the official long-range outlooks issued from mid-Autumn 1998 through Spring 2000.
|
| |
|
Technical Details
|
| |
| In determining these trends, a database of areally-averaged monthly average temperature and total precipitation for 102 climate regions of approximately equal area covering the lower 48 states was used. Each climate region is composed of one or more of the 344 climate divisions outlined on the maps. The monthly data were assembled into time series for each of the 12 3-month periods, as well as for the annual average. The analyses are based on 1941-2005 data for the temperature products, and 1931-2005 for precipitation. Using an approximation based on the results of Livezey and Smith (1999), a best-fit method was used for each climate region and time period, to determine a flat baseline value for the years prior to the mid-1970's, and the linear increase or decrease with time from that baseline value through the ensuing years (up to 2005). It is this best-fit rate of change per unit time since the mid-1970's that is described in these products as the "trend" (expressed in units per decade) for a given region and time of year.
|
| |
For more information, contact Richard Tinker
Climate Prediction Center
at 301-763-8000x7568 or
Rich.Tinker@noaa.gov.
|
|
www.nws.noaa.gov 
