CLDAS土壤相对湿度数据不同时空尺度适用性评估

Evaluation of relative soil moisture from CMA Land Data Assimilation System at different spatiotemporal scales in China

  • 摘要: 基于2020—2021年中国气象局陆面数据同化系统(CLDAS)模拟的逐日土壤相对湿度和土壤水分自动站观测的逐小时土壤相对湿度资料, 采用多个统计对比指标, 在日和月时间尺度、点和区域空间尺度系统性地评估了CLDAS模拟的土壤相对湿度适用性。结果表明: CLDAS模拟的土壤相对湿度与观测值具有一致的逐日变化规律; 0~10 cm、0~20 cm层次CLDAS模拟值与观测值较为接近, 0~50 cm层次模拟值普遍低于观测值; 各层次上, CLDAS模拟值与观测值的相关系数普遍大于0.6, 均方根误差普遍小于30%。区域尺度上, 0~10 cm层次CLDAS模拟值和观测值相关系数为0.78~0.95, 以华南最高; 均方根误差为5.70%~17.26%, 以华东最小; 偏差为−6.63%~15.80%, 以华中偏差绝对值最小。0~20 cm层次CLDAS模拟值和观测值相关系数为0.78~0.95, 以东北和内蒙古最高; 均方根误差为4.45%~14.03%, 以华东最小; 偏差为−5.36%~12.56%, 以华东偏差绝对值最小。0~50 cm层次CLDAS模拟值和观测值相关系数为0.68~0.97, 以东北最高; 均方根误差为4.00%~15.83%, 以华东最小; 偏差为−9.83%~9.62%, 以东北偏差绝对值最小。区域月尺度CLDAS模拟值和观测值整体的相关性较好, 相关系数在6—10月较高; 均方根误差普遍小于15%, 华东各个层次逐月均方根误差均较小。总体来看, CLDAS模拟的土壤相对湿度空间上在华东、华中和东北精度最高, 时间上在6—10月精度最高。

     

    Abstract: Based on daily relative soil moisture from the CMA Land Data Assimilation System (CLDAS) and hourly relative soil moisture from automatic soil moisture stations during 2020−2021, CLDAS relative soil moisture was evaluated in terms of accuracy and suitability by using multiple statistic indices at temporal scales of day and month as well as spatial scales of station and region. The results showed consistent daily variation in CLDAS relative soil moisture and observed relative soil moisture. CLDAS relative soil moisture at depths of 0−10 cm and 0−20 cm was close to the observed relative soil moisture, whereas CLDAS relative soil moisture at a depth of 0−50 cm was smaller than the observed relative soil moisture. The correlation coefficients between CLDAS relative soil moisture and observed relative soil moisture at the three depths were generally greater than 0.6, between which the root mean square errors (RMSE) were smaller than 30%. At the regional scale, the correlation coefficient between CLDAS relative soil moisture and observed relative soil moisture at a depth of 0−10 cm was 0.78−0.95, with the largest value in South China. The RMSE between CLDAS relative soil moisture and observed relative soil moisture at a depth of 0−10 cm was 5.70%−17.26%, with the smallest value in East China. The bias between CLDAS relative soil moisture and observed relative soil moisture at a depth of 0−10 cm was −6.63%−15.80%, with the smallest absolute value in Central China. At a depth of 0−20 cm, the correlation coefficient between CLDAS relative soil moisture and observed relative soil moisture was 0.78−0.95, with the largest value in Northeast China and Inner Mongolia. The RMSE between CLDAS relative soil moisture and observed relative soil moisture at a depth of 0−20 cm was 4.45%−14.03%, with the smallest value in East China. The bias between CLDAS relative soil moisture and observed relative soil moisture at a depth of 0−20 cm was −5.36%−12.56%, with the smallest absolute value in East China. At a depth of 0−50 cm, the correlation coefficient between CLDAS relative soil moisture and observed relative soil moisture was 0.68−0.97, with the largest value in Northeast China. The RMSE between CLDAS relative soil moisture and observed relative soil moisture at a depth of 0−50 cm was 4.00%−15.83%, with the smallest value in East China. The bias between CLDAS relative soil moisture and observed relative soil moisture at a depth of 0−50 cm was −9.83%−9.62%, with the smallest absolute value in Northeast China. The correlation coefficients between monthly CLDAS relative soil moisture and observed relative soil moisture were generally high, with larger values between June and October. The RMSE between monthly CLDAS relative soil moisture and observed relative soil moisture was smaller than 15%, with the smallest value in East China. Overall, CLDAS relative soil moisture performed well in East China, Central China, and Northeast China at the spatial scale and during June−October at the temporal scale.

     

/

返回文章
返回