GEE-based Near-real-time,Medium And High Spatial Resolution Forest Fire Monitoring In Inner Mongolia Autonomous Region

Forest resources are one of the most important natural resources of mankind.Not only maintaining the dynamic balance of the entire nature,it also continuously improves the living environment of human beings.Forest fires refer to those forest fires that lose man-made control,spread and expand freely in woodlands,and cause certain damage and loss to forests and humans.Forest fires are at the top of the three natural disasters(disease,insect pests,and fires)that destroy forests.They are very sudden,destructive,and difficult to deal with and rescue.In recent years,with the impact of climate warming and human factors,the frequent occurrence of forest fires has become one of the most serious forestry disasters that harm forest resources,which not only causes huge losses to human economic construction,destroys the ecological environment,but also threatens the safety of people’s lives and property.Timely and accurate monitoring of forest fires is of great significance to the protection of forest resources and the safety of people’s lives and property.Remote sensing satellites have special capabilities and potential in forest fire point monitoring with wide coverage and high spatial and temporal resolution.Traditional remote sensing-based forest fire monitoring is mainly based on low spatial resolution data,such as MODIS and NOAA-AVHRR.Although the monitoring timeliness is good,it is not possible to detect small-scale forest fire points,which makes it impossible to quickly grasp the fire.In recent years,with the rapid development of medium-and high-resolution satellite observation,Landsat 8 and Sentinel 2 has provided the possibility to solve this problem.This article uses the Google Earth Engine platform,selecting Landsat 8 and Sentinel 2 medium and high spatial resolution data as data source,carry out near real-time medium and high spatial resolution forest fire point monitoring through utilizing short-wave infrared and near infrared channel characteristics.We focus on massive data preprocessing,forest Fire point detection algorithm optimization,fire point misjudgment removal and other key technologies,and finally realize the monitoring of forest fire points with medium and high spatial resolution in Inner Mongolia Autonomous Region within 48 hours.The results show that the middle and high spatial resolution forest fire points detected in this study have high accuracy,and combined with traditional forest fire monitoring,it can provide valuable information for national forest fire monitoring more accurately and efficiently.