摘要
以EOS/MODIS为数据源,采用地表温度（LST）遥感信息模型耦合土壤温度CERES模型反演多层裸土土温。结果表明：土壤温度的模拟值与实测值的变化趋势比较一致,不同土层反演值和地面观察值的相关性均达极显著水平;5～60 cm土层土壤温度的平均绝对误差随土层加深逐渐降低,40～60 cm土层土壤温度的平均绝对误差降至1℃以内,但80、160 cm两层的反演精度误差则相对较大,说明结合其他模型,遥感能够反演一定深度土层的土壤温度,并达到一定精度。
The temperature of different bare soil layers in Shaanxi was retrieved with split windows method of LST coupled with CERES Model based on EOS/MODIS images.The results showed that the change trend of predicted values and measured values of soil temperatures were relatively consistent.The relationship between the soil temperatures retrieved and the measured reached extremely significant level.The average absolute errors decreased from 5 cm to 60 cm deep in the soil.It was below 1℃ at 40 and 60 cm soil layers,but relatively higher at 80 and 160 cm layers.Therefore,the temperature of some deep soil layers could be retrieved by remote sensing coupled with other models to some extent in precision.
出处
《安徽农业科学》
CAS
北大核心
2008年第20期8849-8851,共3页
Journal of Anhui Agricultural Sciences
基金
国家重点基础研究发展计划(973计划)资助(2005CB121108)
江苏省"青蓝工程"资助
南京信息工程大学重点科研基金资助
关键词
地表温度(LST)
土壤温度
CERES模型
Land surface temperature（LST）
Soil temperature
Crop Environment Resource Synthesis（CERES） Model
