Premium
Spatial interpolation of precipitation in a dense gauge network for monsoon storm events in the southwestern United States
Author(s) -
Garcia Matthew,
PetersLidard Christa D.,
Goodrich David C.
Publication year - 2008
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/2006wr005788
Subject(s) - rain gauge , interpolation (computer graphics) , meteorology , precipitation , multivariate interpolation , spatial variability , monsoon , environmental science , cluster analysis , spatial dependence , mathematics , computer science , geography , statistics , artificial intelligence , bilinear interpolation , motion (physics)
Inaccuracy in spatially distributed precipitation fields can contribute significantly to the uncertainty of hydrological states and fluxes estimated from land surface models. This paper examines the results of selected interpolation methods for both convective and mixed/stratiform events that occurred during the North American monsoon season over a dense gauge network at the U.S. Department of Agriculture Agricultural Research Service Walnut Gulch Experimental Watershed in the southwestern United States. The spatial coefficient of variation for the precipitation field is employed as an indicator of event morphology, and a gauge clustering factor CF is formulated as a new, scale‐independent measure of network organization. We consider that CF < 0 (a more distributed gauge network) will produce interpolation errors by reduced resolution of the precipitation field and that CF > 0 (clustering in the gauge network) will produce errors because of reduced areal representation of the precipitation field. Spatial interpolation is performed using both inverse‐distance‐weighted (IDW) and multiquadric‐biharmonic (MQB) methods. We employ ensembles of randomly selected network subsets for the statistical evaluation of interpolation errors in comparison with the observed precipitation. The magnitude of interpolation errors and differences in accuracy between interpolation methods depend on both the density and the geometrical organization of the gauge network. Generally, MQB methods outperform IDW methods in terms of interpolation accuracy under all conditions, but it is found that the order of the IDW method is important to the results and may, under some conditions, be just as accurate as the MQB method. In almost all results it is demonstrated that the inverse‐distance‐squared method for spatial interpolation, commonly employed in operational analyses and for engineering assessments, is inferior to the ID‐cubed method, which is also more computationally efficient than the MQB method in studies of large networks.