Study on NDVI changes in Weihe Watershed based on CA–Markov model

Abstract

The simulation and prediction of the NDVI are conducive to regional ecological restoration. In this research, the Weihe Watershed is selected as the study area, which is located on the eastern edge of the Tianshan tectonic belt. Firstly, according to NDVI characters, we divide the values into five grades, including very low (<0.3), lower (0.3–0.4), medium (0.4–0.5), higher (0.5–0.6), and very high (>0.6), and analyse the NDVI tempo‐spatial variations in Weihe Watershed from 2006 to 2015. Then, we use CA–Markov model with parameters adjustment to simulate and predict the NDVI distributions in 2015 and 2020, respectively. The results show that (a) at temporal scales, the NDVI in Weihe Watershed increase year by year from 2006 to 2015, indicating a better vegetation recovery. At spatial scales, the NDVI increase from north‐west to south‐east gradually, which is related with the main climate factors such as temperature and precipitation. (b) By adjusting the parameters of state transition suitability atlas in the Markov model, we get the model with higher accuracy to simulate NDVI changes. And the Kappa coefficient between the simulation results and the true values is 0.6123. (c) According to the predictions in 2020, the regions with very high NDVI and higher NDVI mainly locate in Qinling Mountains, Guanzhong Plain, Ziwu Mountains, Jinghe Valley, and Luohe Valley. And these regional area percentages occupying the whole watershed increased by 13.62% and 2.64%, respectively, compared with the area percentages in 2015. The regions with medium NDVI, lower NDVI, and very low NDVI mainly distribute at the north Loess Plateau of the watershed. And these regional area percentages occupying the whole watershed decreased by 7.40%, 6.59%, and 2.28%, respectively, compared with the area percentages in 2015. The above results illustrate that the ecological environment of Weihe Watershed is improving, which can provide scientific reference for assessing the regional ecological risk and formulating ecological protection measures.