Study On The Model Of Light Use Efficiency Based On Multi-Angle Hyperspectral Observation And MODIS Image

Terrestrial ecosystems absorb CO₂ by photosynthesis,and researching photosynthesis is of great importance in understanding carbon balance and regional and global climate change and its impact.Light use efficiency as a key parameter of vegetation photosynthesis has become a key factor in monitoring vegetation primary productivity.With the development of hyperspectral,the photochemical vegetation index?PRI?has shown great potential for the estimation of light use efficiency?LUE?,but the PRI-LUE has a lot of uncertainty.The influencing factors of PRI-LUE were mainly focused on single forest vegetation in the canopy or MODIS data on the landscape scale.but on the site scale,the study of the single vegetation maize farmland ecosystem and the multiple plant mixed urban greenland ecosystem is also less.In this paper,the automatic multi-angle hyperspectral remote sensing platform system based on flux tower is used to measure the high spatial and temporal resolution of vegetation canopy hyperspectral data and digital photo data,combined with flux observation data,around farmland ecosystem and urban greenland ecosystem,finally,analyze the daily and seasonal variation characteristics of carbon flux,PRI and LUE of the two ecosystems,and the effects of different time and space scale,canopy light conditions and vegetation growth period.In order to improve inversion accuracy of LUE,the main factors influencing the inversion of LUE are analyzed.On the one hand,it is beneficial to the simulation of GPP on the regional or landscape scale,and on the other hand,it is helpful to understand the dynamic process of the carbon cycle of the ecosystem and reduce the uncertainty in the carbon cycle and carbon budget.The main conclusions of this study are as follows:?1?Because the ecosystems studied are basically deciduous vegetation,the vegetation growth rate of GPP and NEE is different in different growth stages,whether it is farmland ecosystem or urban greenland ecosystem.During the vigorous growth period,the diurnal variation of GPP and NEE was drastic.In the period of vegetation aging,the change is more gentle.This is because the photosynthesis and respiration are not the same in different growing seasons.?2?PRI varies drastically with different angles of observation and the angle of the sun,but in both ecosystems the changes are also very different.The difference of PRI in the observation direction of the single vegetation type farmland ecosystem mainly comes from the different parts of the observed maize,while the difference of PRI in the urban green space ecosystem is mainly due to the observed different vegetation and parts.The results showed that the responses of different vegetation to PRI-LUE were different,which led to the difference of PRI at different angles and PRI-LUE in different ecosystems.?3?In the analysis of the correlation between PRI and LUE and meteorological factors,it was found that PRI and LUE had the best correlation with soil temperature and photosynthetically active radiation in farmland ecosystem.In urban greenland ecosystem,PRI and LUE were most correlated with photosynthetically active radiation,and there was little correlation between PRI and soil temperature.Therefore,in the ecosystems with different vegetation types,there are great differences between the main environmental factors that affect PRI and LUE due to environmental conditions.?4?On the site scale,based on the hyperspectxal reflection data obtained from the multi-angle hyperspectral platform observation system,the PRI obtained by the RossThick-LiSparse BRDF model and the LUE inversion with the flux data are different in the two ecosystems phenological and temporal and spatial scales have a good correlation,but the optimal model of the two differences.?5?Using MODIS-PRI₆₆₈,light use efficiency inversion model was established.In the urban greenland ecosystem and Daman SuperStation corn farm ecosystem are:PRI₆₆₈=0.7143*LUE+0.0061?R²=0.5107?,PRI₆₆₈=5.3653*LUE+0.0134?R²=0.646?...