A method and apparatus for downscaling chlorophyll products based on MODIS

Abstract

A MODIS-based chlorophyll product downscale method of that invention comprises the follow steps of: a, acquiring MODIS remote sensing reflectance data, Landsat atmospheric top reflectance data and corresponding MODIS sea surface chlorophyll product, and converting the Landsat atmospheric top reflectance data into remote sensing reflectance data; Step b: use U.The STFM model predicts the high-resolution Rrs data in the corresponding blue and green bands. C, using regression model to establish the relationship between Rrs blue and green bands and MODIS chlorophyll; Step d, converting the high-resolution Rrs data into the chlorophyll product data by using the relationship obtained by the regression model. The MODIS-based chlorophyll product downscaling method and device of the embodiment of the invention break through the shortcoming that the traditional image superresolution method cannot be too large for downscaling ratio, and are suitable for conversion between different-scale remote sensing data.

Description

technical field [0001] The invention relates to the technical field of spatial data downscaling, in particular to a method and device for downscaling chlorophyll products based on MODIS. Background technique [0002] The existing global daily MODIS chlorophyll products provided by NASA have a spatial resolution of 1km. Because the spatial resolution is too coarse, it cannot meet the water environment monitoring in coastal areas. The 30m resolution data provided by the Landsat satellite has a higher spatial resolution, but since the revisit cycle is once every 16 days, it cannot meet the frequency requirements for rapid monitoring of the water environment in coastal areas. [0003] For spatial data downscaling methods, there are mainly two types of methods in the world. The first type of method is derived from the image super-resolution technology in the computer field; the second type is from the image fusion technology. [0004] For image super-resolution technology, the c...

AI Technical Summary

Problems solved by technology

First, due to the characteristics of MODIS sensor sweep imaging, the point spread of the entire image is more complicated, and it is not possible to model the imaging process of each pixel with a single point spread function

The second problem is that in the field of image super-resolution, the common downscaling ratio is generally 2 times or 4 times

There are two main problems when the U-STFM model directly drinks the MODIS sea surface chlorophyll products: 1) The U-STFM model requires that the change rates of the MODIS data and Landsat data be consistent in the time series, but due to the difference in atmospheric correction and observation time , this assumption is not guaranteed for leaf MODIS and Landsat chlorophyll products

Therefore, it is impossible to directly use the U-STFM model to perform spatial data size manipulation for chlorophyll products; 2) The U-STFM model segments the changing regions in each time series. In the process of segmentation, in order to make the final fused image have Richer details, so the more detailed the segmentation is required, the better, but when the segmentation is small, it will cause the linear unmixing equation to have no solution and singular solutions

Images