中国北方作物和草地LAI全球产品的精度评价

Evaluation the LAI Precision of Global Products of Crops and Grasses over North China

叶面积指数(Leaf area index, LAI)是陆地表层的生态系统、农业生产、水热能量平衡及全球气候变化等等模型中的关键参量之一。目前已有多种的全球LAI遥感产品,科学评价和对比这些LAI产品精度是准确使用这些数据的重要基础。本研究基于多个时期的地面实测LAI数据和Landsat TM、HJCCD影像数据,对中国北方重要的植被类型玉米、冬小麦和草地的GLOBMAP的LAI数据精度进行验证评价,并将其与MODIS、GEOV1产品进行比较。研究结果表明,在玉米的整个生长季过程中,与MODIS、GEOV1 LAI产品相比较,GLOBMAP LAI产品值与地面实测值、TM影像估算LAI值都更为接近。冬小麦GLOBMAP LAI产品明显低于实际值、HJCCD估算LAI值,冬小麦GLOBMAP LAI产品值较实测值低约2~5左右。草地GLOBMAP LAI值精度不高,其与实测值拟合R^2仅为0.231,比HJCCD影像估算值低0.6~2左右。三种全球LAI产品比较而言,GLOBMAP与MODIS LAI产品值相对较为接近,两者都明显低于GEOV1 LAI产品值,但三种LAI产品都较好刻画了玉米和冬小麦冠层LAI的生长变化过程。本研究可为LAI全球遥感产品准确应用、产品算法改进、精度提高提供科学参考。

Leaf area index(LAI)is a critical variable in land surface ecosystem,agricultural production,water and heat energy balance model and global climate change simulation models.Global LAI product retrieved from remote sensing image is an important LAI data source,it is very necessary to evaluate and asses the global LAI product accuracy for its scientific utilization.Based on the ground measured LAI data of several years and Landsat TM,Environmental Remote Sensing Satellite images(HJCCD),MODIS and GEOV1 LAI data,this study evaluated the GLOBMAP product accuracy of most important vegetation types of corn,winter wheat and grass over North and Northern China.The studied results showed that,during the entire maize growing period,maize GLOBMAP LAI data are closer to the ground measured LAI and estimated LAI from TM images,compared to MODIS and GEOV1 LAI data.For winter wheat,GLOBMAP LAI product was significantly smaller than the LAI data estimated from HJCCD images,they are smaller than the ground measured LAI about 2-5.The grass GLOBMAP LAI data is not accurate enough,the regression R2 of which with ground LAI is only 0.231,most GLOBMAP LAI data are smaller than the estimated data with HJCCD about 0.6-2.In general,although the three LAI global products could describe well the general growth process of crop and grass canopy LAI changes,the GLOBMAP LAI and MODIS LAI product values are relatively close,both of which are significantly lower than the GEOV1 LAI product value.This study can provide scientific reference to accurately utilization,algorithm and data improvement of global LAI products.