Bias characteristics of land surface model (CLM4.5) over the Tibetan Plateau during soil freezing-thawing period and its causes
LI Shiyue;YANG Kai;WANG Chenghai;College of Atmospheric Sciences,Lanzhou University/Key Laboratory of Arid Climatic Change and Disaster Reduction of Gansu Province;
The offline simulation experiments over the Tibetan Plateau was conducted through CLM4. 5(Community Land Model version 4. 5) forced by China Meteorological Forcing Dataset. The observations in three sites(D66,TTH and Maqu) were chosen. The CLM4. 5 simulation was compared to the observation and GLDAS(Global Land Data Assimilation System)-CLM2 simulation to analyze the biases of land surface model in simulating soil temperature and moisture during freezing-thawing process and its possible causes. The results showed that simulated soil temperature of CLM4. 5 generally agreed with the observation(averaged RMSE ≈3 ℃),while soil temperature of GLDAS-CLM2 was higher than the observation with the biases(averaged RMSE 6 ℃) larger than that in CLM4. 5,especially in freezing-thawing period; CLM4. 5 can reproduce the seasonal variation of soil moisture,but content of soil moisture had some differences from the observation(averaged RMSE≈0. 1 mm3·mm-3); GLDAS-CLM2 cannot reproduce the seasonal variation of soil moisture. The biases of CLM4. 5 simulation mainly come from the uncertainties of atmospheric forcing,while the biases of GLDAS-CLM2 mainly come from the imperfection of frozen soil parameterizations,besides the uncertainties of atmospheric forcing. Air temperature and precipitation in atmospheric forcing have different effects on the performance of land surface model in freezing-thawing period and non-freezing-thawing period. In non-freezing-thawing period,simulation of soil temperature is mainly affected by the air temperature(r 0. 6),especially before freezing period,the contribution of air temperature biases to soil temperature biases is larger than 50%,the variation of soil moisture is mainly influenced by precipitation,the contribution of precipitation biases to soil moisture biases is about 20%-40%. In freezing-thawing period,soil water flowand heat flux are coupled,the effects of air temperature and precipitation on soil temperature and moisture have reduced,the variation of soil moisture is significantly influenced by air temperature; contribution of air temperature biases to soil moisture biases is 10%-20%. The imperfection of freezing-thawing parameterization schemes in land surface model is the main factor for simulation biases of soil temperature and moisture during freezing-thawing; its effects are larger than the impacts of atmospheric forcing.
【CateGory Index】： S152
【CateGory Index】： S152