Submitted to: Water Resources Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/26/2009
Publication Date: 3/31/2009
Citation: Flerchinger, G.N., Xiao, W., Marks, D.G., Sauer, T.J., Yu, Q. 2009. Comparision of Algorithms for Incoming Atmospheric Long-Wave Radiation. Water Resources Research, 45. W03423. [doi:10/1029/2008WR007394]
Interpretive Summary: Thermal (long-wave) radiation from the atmosphere is an important source of heat for land and water surfaces, influencing surface temperatures, evapotranspiration, snowmelt, and frost, among other important processes. Because thermal radiation is expensive and difficult to measure, data are rarely available and most computer models of terrestrial ecosystems rely on simplifications to estimate atmospheric thermal radiation. Numerous equations are available to do so, but there are virtually no studies that compare how well these equations do under variable atmospheric conditions at more than two or three sites. This study compared 26 different approaches to estimate atmospheric thermal radiation at 14 sites across North America and China. Three algorithms for clear sky conditions and three algorithms for cloudy conditions were found to be superior to most other equation tested. This information can be used to improve predictions from computer models of terrestrial ecosystems.
Technical Abstract: While numerous algorithms exist for predicting incident atmospheric long-wave radiation under clear (Lclr) and cloudy skies, only a handful of comparisons have been published to assess the accuracy of the different algorithms. Virtually no comparisons have been made for both clear and cloudy skies across multiple sites. This study evaluates the accuracy of twelve algorithms for predicting incident long-wave radiation under clear skies, ten cloud correction algorithms, and four algorithms for all-sky conditions using data from twelve sites across North America and China. Data from five sites were combined with publicly available data from nine sites in the AmeriFlux network. Clear sky algorithms that excelled in predicting Lclr were the Dilley-O’Brien, Prata, and Ångström algorithms. Root mean square difference (RMSD) between predicted and measured Lclr averaged 22 to 23 Wm-2 for these three algorithms across all sites. Cloud-correction algorithms of Kimball, Unsworth-Monteith and Crawford described the data best when combined with the Dilley clear-sky algorithm. Average RMSD across all sites for these three cloud corrections was 24 to 25 Wm-2. The Kimball and Unsworth-Monteith cloud corrections require an estimate of cloud cover while the Crawford algorithm corrects for cloud cover directly from measured solar radiation. Optimum limits in the clearness index, defined as the ratio of observed to theoretical terrestrial solar radiation, for complete cloud cover and clear skies were suggested for the Kimball and Unsworth-Monteith algorithms. Based on the results, the recommended algorithms can be applied with reasonable accuracy for a wide range of climates, elevations and latitudes.