The Validation Of The Remote Sensing Albedo Products Based On The LAPVAS

Land surface albedo is a significant parameter in the Earth's energy balance and plays a crucial role in the global change and climate systems.There are many global or regional land surface albedo products with various spatial resolutions and temporal frequencies have been released.The research work about the quality analysis of albedo products becomes more and more important.In order to gain the uncertainty and accuracy of the land surface quantitative products and to provide the basis for wide applications,the LAPVAS was developed based on the theory,methods and practices of the validations.It's validating data sources are the synchronous data from the sites,points and networks of the ground observations.After different kinds of operation for the products data,such as data conversion,scale transform and validating progress,this system can acquire the accuracy of the products.Based on the LAPVAS system,three kinds of albedo products(MuSyQ,GLASS,MODIS)at home and abroad were tested for authenticity.In the process of specific verification,according to the spatial representation of the ground site,respectively,direct verification,multi-scale verification,cross validation and other methods were carried out.On the purpose of investigating the statistical index of albedo products under different surface coverage type classification,the products and in situ data were classified further.For MuSyQ albedo product,the results show that MuSyQ has a high level of accuracy under the condition of majority vegetation cover types.However,in the desert site,whether it is compared with Landsat ETM + aggregation albedo or MODIS albedo product,the results show a greater deviation.Simultaneous analysis of MuSyQ albedo product quality control flag(QC)distribution,the same accuracy problem can also be found.Therefore,the MuSyQ albedo product can reflect the true change of the albedo in the vegetation-covered area,but the data in the desert area need to be further optimized and improved.Using the same reference data and validation methods in the dispose progress of the MuSyQ albedo,the GLASS was validated under the support of the LAPVAS.From the results,it can be found that GLASS has an extremely high level of accuracy as well as a good performance,compared with different reference data.Although in the classification of statistics,with the surface coverage type changes in the accuracy of the situation,But on the whole,GLASS albedo has good quality level under all surface coverage types,and can fully reflect the change of surface albedo.Compared this two kind of albedo products based on the same reference data,the results shows that GLASS has a higher quality than MuSyQ product both at overall and under different land cover type conditions.Finally,based on the verification system and ground data support,the MODIS two different versions(MCD43A3 V005 and V006)albedo products were compared and analyzed,and the accuracy and characteristics of the two albedo products were evaluated.As can be seen from the results,the new version of product not only has a higher temporal resolution,but also a higher level accuracy than old version.However,from the analysis of the distribution of quality control flag results,V006 high-quality data ratio is less than V005 but the proportion of filled data is higher than V005 version.It is obviously that while the improved inversion algorithm improves the temporal resolution lead to the decrease of the high-quality data ratio of V006,but the overall accuracy and the accuracy under different land cover types of the V006 version albedo product is higher than old version,whatever compared with the old data or the ground data.The accuracy of each vegetation type is relatively high relative to V005,and remain stable.MODIS V006 albedo products improving the temporal resolution while maintaining high quality,will play an important role in all relevant research areas.