HR: 17:15h
AN: H24D-06    [Abstracts]
TI: Error sources in the real-time NLDAS incident surface solar radiation and an evaluation against field observations and the NARR
AU: * Park, G
EM: gihyeonp@uci.edu
AF: Department of Civil and Environmental Engineering University of California, Irvine, E4130 Engineering Gateway, Irvine, CA 92697 United States
AU: Gao, X
EM: gaox@uci.edu
AF: Department of Civil and Environmental Engineering University of California, Irvine, E4130 Engineering Gateway, Irvine, CA 92697 United States
AU: Sorooshian, S
EM: soroosh@uci.edu
AF: Department of Civil and Environmental Engineering University of California, Irvine, E4130 Engineering Gateway, Irvine, CA 92697 United States
AB: The atmospheric model is sensitive to the land surface interactions and its coupling with Land surface Models (LSMs) leads to a better ability to forecast weather under extreme climate conditions, such as droughts and floods (Atlas et al. 1993; Beljaars et al. 1996). However, it is still questionable how accurately the surface exchanges can be simulated using LSMs, since terrestrial properties and processes have high variability and heterogeneity. Examinations with long-term and multi-site surface observations including both remotely sensed and ground observations are highly needed to make an objective evaluation on the effectiveness and uncertainty of LSMs at different circumstances. Among several atmospheric forcing required for the offline simulation of LSMs, incident surface solar radiation is one of the most significant components, since it plays a major role in total incoming energy into the land surface. The North American Land Data Assimilation System (NLDAS) and North American Regional Reanalysis (NARR) are two important data sources providing high-resolution surface solar radiation data for the use of research communities. In this study, these data are evaluated against field observations (AmeriFlux) to identify their advantages, deficiencies and sources of errors. The NLDAS incident solar radiation shows a pretty good agreement in monthly mean prior to the summer of 2001, while it overestimates after the summer of 2001 and its bias is pretty close to the EDAS. Two main error sources are identified: 1) GOES solar radiation was not used in the NLDAS for several months in 2001 and 2003, and 2) GOES incident solar radiation when available, was positively biased in year 2002. The known snow detection problem is sometimes identified in the NLDAS, since it is inherited from GOES incident solar radiation. The NARR consistently overestimates incident surface solar radiation, which might produce erroneous outputs if used in the LSMs. Further attention is given to the diurnal cycle of NLDAS. Because LSMs are designed to simulate radiation diurnal variation, they are always running at high temporal resolution (30-minutes to several hours). The NLDAS solar radiation data clearly show much smoothened diurnal cycles than observation in cloudy days, which is due to the lower spatial resolution of the NLDAS inherited from the GOES and the EDAS. Taking these problems discovered above into consideration, it is suggested that the NLDAS project work on reducing the systematic biases in the surface solar radiation estimates, so that the users could take full advantage from the data.
DE: 1833 Hydroclimatology
DE: 1840 Hydrometeorology
DE: 1843 Land/atmosphere interactions (1218, 1631, 3322)
DE: 1855 Remote sensing (1640)
SC: Hydrology [H]
MN: Fall Meeting 2005