Evaluation of NASA satellite- and assimilation model-derived long-term daily temperature date over the continental US

Authors: White, Jeffrey; HOOGENBOOM, GERRIT - University Of Georgia; STACKHOUSE, PAUL - National Aeronautics And Space Administration (NASA); HOELL, JAMES - Science Systems, Inc

Submitted to: Agricultural and Forest Meteorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/14/2008
Publication Date: 3/1/2009
Citation: White, J.W., Hoogenboom, G., Stackhouse, P.W., Hoell, J.M. 2009. Evaluation of NASA satellite- and assimilation model-derived long-term daily temperature date over the continental US. Agricultural and Forest Meteorology, 148:1574-1584.

Interpretive Summary: Agricultural research increasingly is expected to provide precise, quantitative information for specific geographic regions. Availability of weather data constrains efforts to provide such information using computer models and related tools. NASA’s Prediction Of Worldwide Energy Resources (NASA/POWER) project provides daily weather data globally. Most variables are available from 1983 onward. The data are considered average values for a 1°grid of latitude and longitude. This is roughly equivalent to a 70 by 50 mile area for US latitudes, so one concern is whether the area covered by a single cell of is too large. Another concern is that since the data are obtained through a complex process of modeling and adjustment to various types of satellite and weather observations, they may have unsuspected bias or errors. We compared NASA/POWER temperature data from 1983 to 2004 for the continental US with data of 855 individual stations from the National Weather Service Cooperative Observer Program (COOP). Besides direct comparisons of daily data, a wheat model was used to compare how the data performed in predicting development of a wheat crop. The comparisons of daily data generally showed good agreement, although maximum temperatures from NASA/POWER averaged 2.4°C cooler than COOP data, and values of minimum temperatures were 1.1°C warmer than the COOP data. Differences in temperature were greater during the winter months. Simulations of flowering time were also similar, but values for winter wheat regions showed better agreement than southern, winter-grown spring wheat regions. In mountainous regions, the differences between the data sources were strongly associated with elevations, which in large part resulted from NASA/POWER data being based on mean elevations over the 1° grid cells. Overall, the NASA/POWER data are a promising source of temperature data for the USA. Furthermore, it appeared likely that the POWER data could be improved by adjusting for elevation effects and reducing the apparent bias with time of year. Ultimately, use of the NASA/POWER data in models and other tools may allow substantial improvement in decisions of farmers and policy-makers, ultimately benefiting consumers and leading to practices that better conserve water and energy.

Technical Abstract: Agricultural research increasingly is expected to provide precise, quantitative information with an explicit geographic coverage. Limited availability of continuous daily meteorological records often constrains efforts to provide such information through integrated use of simulation models, spatial analysis, and related decision support tools. The Prediction Of Worldwide Energy Resources (NASA/POWER) project at the NASA Langley Research Center provides daily data globally for maximum and minimum temperatures and other weather variables on a 1° latitude-longitude grid. The data are assembled from a range of products derived from satellite imagery, ground observations, windsondes, modeling and data assimilation. Daily temperature data from NASA/POWER for 1983 to 2004 for the continental US were compared with data of 855 individual ground stations from the National Weather Service Cooperative Observer Program (COOP). In addition to direct comparisons of temperature data, a wheat (Triticum aestivum L.) simulation model was used to compare predicted time to anthesis using the two data sources. Comparisons of daily maximum temperatures (Tmax) gave an r2-value of .88 (P < .001) and root mean-square error (RMSE) of 4.1°C. For minimum temperature (Tmin), the r2-value was .88 (P <.001) and RMSE, 3.7°C. For the daily mean temperature (Tavg), the r2-value was .91 (P <.001) and RMSE, 3.2°C. Mean values of Tmax, Tmin, and Tavg from NASA/POWER were respectively 2.4°C cooler, 1.1°C warmer, and 0.7°C cooler than the COOP data. Differences in temperature were least during summer months. When data were aggregated over periods of 2 to 30 days, the RMSE declined to below 2.7 d for Tmax and Tmin, indicating that the NASA/POWER data would be more useful in applications requiring data at a weekly rather than daily time scale. Simulations of time to anthesis with the two data sources were also strongly correlated (r2 = .92, P < .001, RMSE = 14.5 d). Anthesis dates of winter wheat regions showed better agreement than southern, winter-grown spring wheat regions. In mountainous regions, the differences between the data sources were strongly associated with differences in elevation, which in large part resulted from NASA/POWER data being based on mean elevations over a 1° grid cells vs. COOP data corresponding to the elevation of specific stations. Additional sources of variation, in regions where the elevation difference was minimal, may be proximity to coastlines, as well as the coupled effects of errors in COOP observations and in the POWER data, and differences in sampling time for the respective maximum and minimum temperature values, although these factors were not quantified. Overall, if mountainous and coastal regions are excluded, the NASA/POWER data appeared promising as a source of continuous daily temperature data for the USA for research and management applications concerned with scales appropriate to the 1° coordinate grid. Furthermore, it appeared likely that the POWER data could be improved by adjusting for elevation (lapse rate) effects, reducing apparent seasonal bias, and better estimation of actual maximum and minimum temperatures in diurnal cycles.