干旱区研究 ›› 2014, Vol. 31 ›› Issue (4): 611-618.

• 天气与气候 • 上一篇    下一篇

对陆面过程模型CoLM中关键参数的修正

  

  1. (1. 新疆大学资源与环境科学学院,新疆 乌鲁木齐 830046; 2. 绿洲生态教育部重点实验室,新疆 乌鲁木齐 830046; 3. 新疆大学智慧城市与环境建模重点实验室,新疆 乌鲁木齐 830046; 4. 中国气象局乌鲁木齐沙漠气象研究所,新疆 乌鲁木齐 830002; 5. 塔克拉玛干沙漠大气环境观测试验站,新疆 塔中 841000)
  • 收稿日期:2013-03-25 修回日期:2013-05-23 出版日期:2014-07-15 发布日期:2014-08-11
  • 作者简介:刘永强(1969-),男,博士,副教授,主要从事陆面过程参数化与模拟研究.E-mail: Lyqxju@163.com
  • 基金资助:

    国家自然科学基金项目(41265002,41175140);科技部公益性(气象)科研专项(GYHY201306066);中国沙漠气象科学研究基金项目(Sqj2009012)

Modification of Key Parameters in CoLM

  • Received:2013-03-25 Revised:2013-05-23 Published:2014-07-15 Online:2014-08-11

摘要: 通过对塔克拉玛干沙漠腹地观测数据的分析和模拟,重新修正了陆面过程模型CoLM中的关键参数值和参数化方案。① 采用涡动相关数据和MoninObukhov理论,对地表动力学粗糙度重新计算;为提高地表比辐射率的精度,采用两种不同的方法进行计算和比较;将地表反照率参数值校正为符合年际变化的参数化方案;重新选择符合沙漠环境的地表热力学粗糙度参数化方案;改变相同的10层土壤导热率为不同值。② 经过参数值的校正与参数化方案的修正,改善了CoLM在沙漠地区对地表通量、地表和土壤温度模拟的缺陷;③ 改善后的CoLM模拟值与观测值的标准误差小于改善前,但不同季节模拟改善效果存在差异,总体上冬季的标准误差小于其他季节。

Abstract: It is one of the challenges in land modeling to modify and validate the land surface models based on the integrated observation in desert. Particularly, the key parameters and parameterization schemes for desert need to be evaluated insitu so as to improve the models. In this paper, the landsurface key parameters and modified parameterization schemes in the common land model (CoLM) were calibrated based on the annual data (from October 1, 2008 to September 30, 2009) observed in the hinterland of the Taklimakan Desert, and some data were ignored and removed due to the instrument malfunction or electronic and physical noise, [WTBX]etc.[WTBZ] First, aerodynamic roughness length was reevaluated based on the data obtained by Eddy Covariance System and MoninObukhov similarity theory; the accuracy of surface emissivity was improved with two different methods, which was calculated by contemporaneous different data; the fixed parameter of surface albedo was revised to parameterization scheme in order to consistent with its annual variation; the scheme of thermal roughness length that performed best was selected to replace the defaulted one; the original values of soil thermal conductivity were equal in 10 layers when the model was run, they were calibrated into different values in different layers. Second, the key parameters and parameterization schemes were revised and simulated in CoLM. The simulated values were compared with the observed ones. Third, the variation of standard deviation between the observation and simulation by revised parameters or parameterization schemes in every month were compared and analyzed.