ObjectiveGrassland fire is a key disturbance factor in natural ecosystems and it greatly influences ecosystem structure and function. XinBarag Prairie of northwestern China is one of the important grassland ecosystems in China. With the frequent occurrence of extreme weather events, more extra serious wildfires will occur in this region. This study could provide a scientific basis for zoning grassland fire risks and wildfire management.

MethodIn this study, the spatial and temporal patterns of wildfires in XinBarag Prairie were analyzed using statistical software, GIS and ENVI based on MODIS data from 2001 to 2016. Through the integration of meteorological data, the impact of meteorological factors on the characteristics of monthly variations in grassland fire was investigated.

ResultXinBarag Prairie had regular distribution in the spatial and temporal patterns of wildfires. In the temporal patterns, annual burned area had a fluctuation period of 10 years generally in the whole region. The monthly burned area had significant differences among various months and the burned area was mainly distributed in April, May, September and October. The time corresponds to spring and autumn, in which fire disaster occurred frequently. In spatial patterns, the high frequencies of wildfires were all widely distributed near the country s borders. The large-scale fires crossing the border line will pose a serious threat to the steppe in China; There were close relations between grassland fire and meteorological factors. Grassland fires easily occurred and spreaded in areas with monthly average temperature ranged between 0 and 10 ℃, monthly total rainfall ranged between 0 and 20 mm, monthly mean relative humidity was below 40% and monthly mean wind speed was above 5 m/s.

ConclusionThe preventing region of grassland fire in XinBarag Prairie should concentrate in its western, southeastern and northeastern regions, and the monitoring and management of grassland fire in border region should be strengthened.In time, we should concern April, May, September and October, especially the areas with monthly average temperature between 0 and 10 ℃, monthly total rainfall between 0 and 20 mm, monthly mean relative humidity below 40% and monthly mean wind speed above 5 m/s. In addition, future research will focus on the fluctuation period of grassland fire from longer temporal and wider spatial scale to explore the dynamic change of wildfires and its influencing factors more effectively.