Evaporation remote sensing inversion application based on reflectivity-vegetation coverage two-dimensional space

Abstract

The invention discloses an evapotranspiration remote sensing inversion application based on a reflectivity-vegetation coverage two-dimensional space, and belongs to the technical field of surface evapotranspiration remote sensing estimation, and the method comprises the steps: S1, constructing an evapotranspiration remote sensing inversion network based on a convolutional neural network; S2, training the evapotranspiration remote sensing inversion network by using the training sample; S3, inputting the meteorological data and the satellite remote sensing image of the target area in the presettime scale into the trained evapotranspiration remote sensing inversion network to obtain evapotranspiration data of the target area in the preset time scale. According to the evapotranspiration remote sensing inversion application based on the reflectivity-vegetation coverage two-dimensional space, the neural network and the method for accurately obtaining the surface evapotranspiration of the target time scale through the remote sensing means are perfectly combined, the evapotranspiration calculation speed is increased, the calculation period is shortened, operation is easy, and intelligenceis good.

Description

technical field [0001] The invention belongs to the technical field of remote sensing estimation of surface evapotranspiration, and in particular relates to an application of remote sensing inversion of evapotranspiration based on the two-dimensional space of albedo-vegetation coverage. Background technique [0002] Evapotranspiration (Evapotransspiration, ET) is an important part of the surface water cycle and energy balance, which determines the water and heat transfer in the soil-vegetation-atmosphere system. , play an important role. Remote sensing technology is considered to be the most effective method to obtain regional scale evapotranspiration distribution on the earth's surface from the aspects of technology, economy and effectiveness. [0003] At present, remote sensing evapotranspiration estimation methods have one-source model, two-source model and surface temperature-vegetation index triangle / trapezoidal model. However, the traditional estimation model has the...

AI Technical Summary

Problems solved by technology

However, the traditional estimation model has the following limitations: (1) It requires matching data of visible light and thermal infrared, which directly excludes the application of satellite remote sensing sensors that have no thermal infrared band but only visible light band; (2) because the surface temperature is not only affected by soil The impact of water content is also affected by meteorological factors such as solar radiation, air temperature, relative humidity, and wind speed. Therefore, it is impossible to carry out the dry edge of the surface temperature-vegetation index triangle / trapezoidal model using only satellite single transit data or long-term series data. And wet edge determination, at the same time, it takes time to calculate and calibrate the parameters every day, and in the calculation process, the server memory occupies a large amount, the server requires high configuration, the calculation cycle is long, the operation is complicated, the professionalism is strong, and the intelligence is poor

Therefore, the application of the calculation method for building models using remote sensing short-wave infrared reflectance data has limitations

Images