Use the NDVI tool in the ENVI 5.3 Toolbox to calculate the band of the images in the study area. The formula is NDVI=(NIR-R)/(NIR+R), where NIR is the near-infrared band and R is the red light band. Select the calculated bands in the NDVI tool and output the results, and finally get the NDVI values for each month within the buffer zone. Use the Maximum Value Composite (MVC) formula: MaxNDVI_i=Max(NDVI_1, NDVI_2, NDVI_3, NDVI_4), where: MaxNDVI_i represents the maximum monthly NDVI value; i is the month, with values ranging from 5 to 10; and [NDVI]_1-[NDVI]_4 respectively represent the NDVI values of raw images after preprocessing in the corresponding month. Obtain the NDVI values during the growing season in the study area, calculate the mean NDVI within a unit distance, and use the grid calculator in ArcGIS10.2.2 to calculate the mean NDVI values for each month in the growing season.

The 1977 image is a first-level data product of the Landsat1-5 Multispectral Scanner (MSS), with a spatial resolution of 60m and a temporal resolution of 18d. Images from 1987, 1997, and 2007 are first-level spectral data products from the Landsat4 and Landsat5 thematic mappers (TM) with a spatial resolution of 30m and a temporal resolution of 16d. The 2017 image is a first-level multispectral data product from andsat8 OLI(Operational Land Imager) and TIRS(Thermal Infrared Sensor), with a spatial resolution of 30m and a temporal resolution of 16d. The spatial resolution of the processed data accuracy is the same as the original data, and the temporal resolution is months and years.