Forest fire behavior potential prediction method based on multi-source data fusion

Abstract

The invention provides a forest fire behavior potential prediction method based on multi-source data fusion, and belongs to the field of forest fire early warning. The forest fire behavior potential prediction method comprises the following steps: identifying combustible types according to airborne hyperspectral remote sensing data, estimating vegetation coverage, and inverting vegetation canopy moisture content; inverting a vegetation vertical structure according to the laser radar point cloud data; obtaining a subcompartment factor according to the forest resource planning survey data; obtaining surface combustible information according to the sample plot survey data; and establishing a combustible attribute data table by using the combustible type, the vegetation coverage, the vertical structure, the subcompartment factor and the ground surface combustible information and taking a subcompartment as a unit, generating three potential indexes and a spatial distribution diagram of a ground surface fire behavior, a crown fire behavior and an effective combustible, and carrying out forest fire behavior potential prediction. According to the method, the forest fire behavior potential is predicted by taking a subcompartment as a unit on the forest farm scale, so that the forest fire behavior potential is predicted more accurately and timely, and prevention and timely fighting of forest fire disasters are realized.

Description

technical field [0001] The invention belongs to the field of forest fire early warning, in particular to a forest fire behavior potential prediction method based on multi-source data fusion. Background technique [0002] Forest fire risk forecasting is generally divided into three types: fire risk weather forecast, forest fire occurrence forecast and forest fire behavior forecast. With the comprehensive application of modern high-tech information technology represented by remote sensing information processing technology, geographic information system technology, computer technology and network technology in forestry, forest and grassland fire early warning has changed from a traditional experience model to a modern one dominated by science and technology. The intensive management mode mainly focuses on how to use modern information technology to simulate the spread speed and direction of the fire scene in the forest fire, but it still lacks the potential for forest fire beha...

AI Technical Summary

Problems solved by technology

Since the forest resources survey and planning data usually need to be updated once every ten years, it is difficult to update the fuel parameters in a large area through manual ground surveys, and it requires a lot of manpower and material resources. Fire behavior prediction needs, existing methods cannot accurately and timely predict forest fire behavior potential

Images