Inversion And Validation Of FPAR Based On Remotesensing Data And Field Sampling

Fraction of Absorbed Photosynthetically Active Radiation(FPAR) is the main vegetation physiological parameter and the important land characteristic parameter of ecosystem, is often used to describe the exchange process between vegetation structure and related material and energy, and is closely related to photosynthesis which is the process of vegetation leaf absorption and the conversion of light energy. At the same time, FPAR is also an effective auxiliary factor to sign the growth of green vegetation and monitor land cover.Hulunber Grassland is one of the most complete preservation grassland in China or even the whole world, and is the most concentrative and representative area of temperate meadow-steppe in China. At present, inversion and validation of FPAR has become a hot issue that many experts studied in it, and has a obvious trend of more attention and concern. This paper based on Hulunber Grassland Ecosystem Observation and Research Station, CAAS, taking meadow-steppe in Sher Tara, Hulunber, Inner Mongolia as example, carried on comparative analysis in inversed FPAR from HJ-1CCD and Landsat-8OLI data in the study area, confirm the applicability of HJ-1CCD high resolution image, and then achieve the validation of MODIS/FPAR product through the bridge of HJ-1CCD images. This study improve the accuracy of FPAR estimation that using Remote Sensing technology, provide a theoretical basis and support for the further study. The main contents and conclusions are as follows:(1) Inversed FPAR using HJ-1CCD and Landsat-8OLI data which has a similar getting time and FPAR in the study grassland which observed by Accu PAR that made in LI-COR company, America, then carried on the thorough careful analysis to the results. The purpose is provide some reference and guidance for the further study, at the same time, promote the application of HJ-1CCD images in this area. The results showed that: there are significant positive correlation between measured FPAR and NDVI calculated from CCD and OLI data; the percentage in the difference between FPAR inversed from CCD data and OLI data respectively which does not exceed 5% is up to 94% in the whole cutting image which including study area; excluding the cloud and cloud shadow region, there is a good correlation between FPAR inversed from CCD data and OLI data, besides R2 achieved to 0.6822.(2) Using HJ-1CCD high resolution images and field FPAR value to achieved the validation of MODIS/FPAR product, the object of study is Hulunber meadow-steppe which is throughout the entire growing season from June to August, in order to form a more rational sampling strategy and improve product accuracy. MODIS/FPAR value is overestimated almost 44.3% than 1km HJ/FPAR “true value” on the scale of 1km×1km.On the plot scale(3km×3km), there is a significant positive correlation between MODIS/FPAR value and HJ/FPAR “true value”, R2 is higher than 0.9, and the trend of variety is also uniformity throughout the whole growing season, in addition has the same overestimation.