Spatiotemporal dynamics of fractional vegetation cover and climate response in Inner Mongolia during 1982-2021 based on GEE

Fractional vegetation cover (FVC) has been one of the most widely utilized indicators to assess the vegetation growth status for the better ecological environment. The spatiotemporal evolution of FVC and its relationship with the climate change can also be crucial to promote the ecological civilization and the sustainable development in agriculture and animal husbandry. However, the large-scale and long-term FVC research is often required to deal with the massive image data, leading to the significant challenges on research advancement. The Google Earth Engine (GEE) platform can be expected to offer the new and effective solutions. In this study, a perfect tradeoff was gained between the rich data sources and powerful processing using GEE platform. A dimidiate pixel model (DPM) was employed to produce the FVC under the different types of land cover in Inner Mongolia, northern China. The Sen's slope estimator and Pearson correlation coefficient were then used to identify the temporal and spatial change of FVC and its correlation with the climate factors. Initially, the FVC annual time series were collected from 1982 to 2021 using the NOAA AVHRR NDVI data within GEE. Subsequently, Sen's slope estimation and the Mann-Kendall test were conducted to analyze the significant trends of FVC over the four decades. Finally, Pearson correlation coefficients were calculated between FVC and climate factors. The significance tests were carried out to reveal the relationship between FVC and climate change. Additionally, the spatial distribution maps were obtained to depict the spatial pattern of the correlations in the significant correlated regions. The results showed that: 1) There was the significant band pattern in the spatial distribution of the average FVC, with the gradual increase from the west to the northeast. A large difference of FVC was also found in different types of land cover. Generally, the FVC was >0.1-0.2 in 18.7% of the region. The proportions of FVC >0.5-0.6 and >0.7-0.8 were both close to 15%. There were the relatively small proportions of 0-0.1 and >0.8-0.9, particularly without the proportions of >0.9-1.0. The rest intervals were about 10%. 2) The FVC decreased significantly in the desert area of western Inner Mongolia, accounting for 13.1%. There was no significant trend of FVC in the mountains and grasslands of the middle, where the area was accounted for 38.2%. A very significant upward trend was found in the FVC of croplands, forests and part of grasslands in the south, east and northeast, accounting for 38.3% of the total. 3) There were the significantly different correlations between FVC and meteorological factors in the different types of land cover. The FVC of grassland was significantly or extremely significantly positive correlated with the annual precipitation, while the FVC of croplands was significantly or extremely significantly positive correlated with annual average temperature, and the FVC in some areas of forests was significantly negative correlated with annual precipitation, as well as the FVC in some barren was significantly negative correlated with annual average temperature. A 40-year-long time series of annual FVC were successfully constructed using NOAA AVHRR NDVI data on the GEE platform. The robust processing and analysis capabilities of platform were better balanced to identify the trends of FVC in the various land cover types, together with the correlation with climate change. These findings can provide some significant implications to guide the ecological environment protection for the sustainable development in Inner Mongolia of China.