| 摘要: |
| 陆面模型为区域农田土壤墒情监测提供了很好的途径,优化选择模型的网格尺度可以最有效地的利用空间输入信息,提高计算效率。本研究以海河平原内的1°×1°(115.5~116.5°(E),38~39°(N))为研究区,运用陆面模型CLM3.0分别在(1/120)~1°的14种不同网格尺度上对2003年3—5月的土壤墒情进行了独立模拟,分析在一定精度的空间输入数据条件下,陆面模型的网格尺度在该区域春季土壤墒情模拟中的优化取值。研究表明,结合模型输入数据的空间分辨率选择合适的网格尺度,可有效地减少计算机浮点计算取舍引起的误差;网格的无限精细并不能提高模拟效果,需要依据土壤砂粒百分含量数据的精度、变程及模拟目的优化选择陆面模型的网格尺度。当仅需要获得区域的土壤墒情平均值时,网格尺度的优化取值在土壤砂粒百分含量数据变程的1.4倍附近;当需要获得区域的土壤墒情空间变异特征时,网格尺度的优化取值在土壤砂粒百分含量数据变程的28%附近;当需要获得区域的土壤墒情空间变异特征及极大值时,网格尺度的优化取值在土壤砂粒百分含量数据变程的19%附近;当需要获得区域的土壤墒情的所有空间统计特征时,网格尺度的优化取值在土壤砂粒百分含量数据的空间最小尺度附近。 |
| 关键词: 土壤墒情 陆面模型 网格尺度 春季 优化选择 |
| DOI：10.11841/j.issn.1007-4333.2012.05.027 |
| 投稿时间：2012-03-13 |
| 基金项目:国家重点基础研究发展计划项目(2010CB951102);公益性行业(农业)科研专项(201203031) |
|
| Optimal selection of grid-scale for land surface model in the numerical simulation of spring soil moisture |
|
LIU Xin, REN Li
|
| (College of Resources and Environmental Sciences, China Agricultural University, Beijing 100193, China) |
| Abstract: |
| Land surface model provides a suitable way for regional farmland soil moisture monitoring.Optimizing the grid scale of land surface model can improve computational efficiency by making the most effective use of spatial input information.In this study,we used the 1°×1°region (115.5-116.5°(E),38-39°(N)) in Haihe River Plain as the study area.The community land model (CLM3.0) was used to simulate soil moisture during the spring(March-May) of 2003 over 14 grid scales((1/120)-1°).With input data of a certain spatial precision,the optimized values of grid scale for land surface model were surveyed.The results showed that combination with spatial resolution of model input data,rounding errors in computer's floating-point calculations could be effectively reduced by an appropriate selection of grid scales.Unlimited fine grid scale did not improve modeling accuracy.An optimum grid scale for land surface model could be chosen based on the precision and range of soil sand percentage data,and the purpose of the simulation.The optimal grid scale was about 1.4-times the range of soil sand percentage data for average soil moisture in the region,around 28% of the range of soil sand percentage data for the regional spatial variability of soil moisture,around 19% of the range of soil sand percentage data for the regional spatial variability of soil moisture and the maximum value,and around the minimum spatial scale of soil sand percentage data for all spatial statistical features of soil moisture is required. |
| Key words: soil moisture land surface model grid scale optimal selection |
