Satellite observed rapid green fodder expansion in northeastern Tibetan Plateau from 2010 to 2019

The livestock product consumption per capita in China has almost doubled in the past three decades. The planting of green fodder has increased in the agropastoral ecotone of China to meet the increasing demand for livestock feed, and the green fodder expansion can have subsequent consequences on the environment. However, information on the area and distribution of green fodder is very limited. Here, we developed a pixel- and phenology-based algorithm to map green fodder and track its dynamics in the northeastern Tibetan Plateau, a typical alpine pasture region in China, using all the available Landsat images and Google Earth Engine (GEE). We developed a simple approach for the rapid identification of green fodder fields by using a new green fodder index that considers the unique phenology of green fodder, which has higher greenness and water content in the late growing season than other vegetation. A total of 858 Landsat images were used to generate green fodder maps in northeastern Tibetan Plateau (including Zeku, Guinan, and Tongde Counties) in three periods (circa 2010, 2015, and 2019). The overall accuracies of our green fodder maps were 93-97% and the Matthews correlation coefficients were 0.76-0.83. We found a rapid expansion of green fodder from 16.3 km(2) in 2010 to 136.1 km(2) in 2019. Newly cultivated green fodder occurred in both existing croplands and natural grasslands. Our study demonstrated the potential of the phenology-based approach, all the available Landsat images, and GEE for tracing the historical dynamics of green fodder at 30-m resolution in alpine regions. Our findings advance our understanding of changes in forage area, production, the supply-demand gap, and the ecological and climatic consequences of green fodder expansion.