| Forest fires not only harm a lot of vegetation and change ecological pattern prefire,but also the number of forest burning biomass and carbon emissions are closely related.Exploring the use of long time series of satellite remote sensing data products to reflect the temporal and spatialcharacteristics of forest fires and forest burning biomass for nearly 14 years in domestic forest land,was the major objective of the current work from the perspective of qualitative and quantitative.The national MODIS satellite burned area products(MCD45A1)and active fire detection products(MOD14A2)from 2001 to 2014 datasets has been selected for the study.At first,the feature of power law distribution of FRP(fire radiative power)was analyzed on forest type;Meanwhile,the estimating burned forest biomass models have been built on forest type.The scaling parameter m has been gotton by using the linear regression fit to probability distribution function in log scales.The annual fire duration d has been modified by using the selected forest fire about 10 times every year.The national burned forest biomass was estimated on an annual basis.At the same,the study results calculated from MODIS burned dataset(MCD45A1)were compared with those derived from forest fire emissions calculation model to validate the efficacy of the current estimation method;Secondly,the MCD45A1 data was analyzed by ArcMap10.0 space statistical analysis method and Excel statistical analysis software for evaluating temporal and spatial distribution pattern of forest fires.The sensitive region of domestic forest fires has also been analyzed by the single indicator and composite indicator.This research results are as follows:(1)The burned forest biomass estimation method has been developed based on power law distribution modelAccording to the FRPs of broadleaf forest,coniferous forest and shrub forest following the power law distribution by analysising the FRP of the MODIS data fire product sets for the nearly 14 years,the estimation models for three forest types have been estimated to calculate the total burned forest biomass in year 2001-2014 at national scale.In 14 years,Average annual burned biomass of broadleaf forest was in 0.94Tg-1.37 Tg,with a burned biomass in 0.80Tg-1.92 Tg for coniferous forests,shrub forest’s burns in 0.37Tg-0.53 Tg.The results of the two methods were significantly difference in some years,even certain values derived from forest fire emissions model were far beyond the total burned biomass of the observed 14 years calculated from the method developed in the current work.The results and inter annual variation were more consistent with the statistics of fire numbers and burned area issued by the national Bureau of Statistics.(2)The national evaluation method of temporal and spatial distribution patterns and regional sensitivity of forest fires have been formed using remote sensingThe temporal and spatial distribution pattern and the sensitive region evaluation for forest fires has been analyzed using MCD45A1 burned datasets and other statistics datas in 2001-2014 years.Firstly,the temporal distribution of forest burned area and the charts of forest burned area and other thematic maps in 2001-2014 years at the national scale has been studyed.At the same times,the characteristics of domestic seasonally forest fires were analyzed for nearly 14 years.Namely spring and autumn in the burned area accounted for 84% of the four seasons’ total area,and the burned area in fall is 40% higher than in the spring,it was explained that the preventive measures and the public consciousness during spring are better than fall,and it showed that the autumn also strengthened the awareness of forest fire prevention and management efforts.Secondly,the three kinds of single indicator and composite indicator have been taken to evaluate regional sensitivity of national forest fire.The results of three kinds of single index calculation,for example the forests burn rate,the proportion of forest burned area and coefficient variation of forest burned area,was very significant difference.Based on the results of the single index,the equation to solve thecomposite indicator values(CI)for provinces(cities)has been developed.The 31 CI values as 31 samples,the T test in 99% reliability has been analysed.The average confidence interval of CI is from 13.06 to 18.94.The value of comprehensive index falls in this interval as "moderate" sensitive areas,of which there are 10 provinces(cities)fall within this range;the value of CI of 11 provinces(cities)is higher than the average value,they were considered "light" sensitive area;there also were 10 provinces(cities)area,which CI is lower than the average value,so they has been classified as "serious" sensitive areas. |
PDF Full Text Request