| China is a large country of grasslands,with 400 million hectares of various types of natural grasslands,accounting for 41.7% of the country's land area,ranking second in the world.The grassland is not only a production base for animal husbandry,but also the basis for the protection of ecological security barriers and the inheritance of pastoralist life and grassland culture.It has living,productive and ecological functions.Fire is one of the environmental factors of grassland ecosystems.Fire is widely used in grassland pastures and is also used to increase farm productivity.At the same time,in the international,China is also a country with severe grassland fires.Among the 400 million hectares of grassland,the relatively high frequency of grassland fires is 677 million hectares,and the grassland fires are prone to 130 million hectares.Grassland fire is a kind of natural disaster.It has strong destructiveness,strong suddenness,and difficulty in disposal and rescue.Once it occurs,it will have a huge impact on the ecological environment of the pastoral area and the safety of people's lives and property.Therefore,scholars at home and abroad have increasingly studied grassland fires and the factors that affect grassland fires,and have also achieved certain results.Grassland fire is one of the most important interfering factors in the natural ecosystem.Since grassland fires must occur under certain climatic backgrounds and meteorological conditions,climatic factors and meteorological factors are closely related to the occurrence and development of grassland fires.The current grassland fire weather index model only reflects the impact of short-term meteorological factors on grassland fires,and there should also be a mid-and long-term climate scale impact on grassland fires.This paper takes Hulun Buir as the study area,and analyzes the impact of the inter-month climatic factors on grassland fire in 2000-2014 using statistical analysis methods,cross-correlation analysis methods,Logistic regression,and spatial analysis.The following results are as follows:(1)High-,medium-,and low-confidence fire spots were extracted from the hot anomaly data of the Moderate-resolution Imaging Spectroradiometer(MODIS)—MOD14A1(Terra)and MYD14A1(Aqua).The land use data were used to screen them and extract the study area for nearly 15 years.Based on the analysis of the grassland fire point in the study area,there were a total of 5565 grassland fires in the study area in the past 15 years.In spatial terms,the number of grassland fires in Oroqen Zizhiq,Molidawa Zizhiqi,Arun Qi,Zalantun Shi,Chen Barag Qi,Manzhouli Shi,Hailar Shi and Ergun Shi is higher than other City.The number of grassland fires in Genhe Shi,Xin Barag Zuoqi,Xin Barag Youqi,Yakeshi Shi,and Ewenkizu Zizhiqi is relatively small.In terms of time,the number of fires in the Hulun Buir grassland in 2000 was the highest,accounting for 913 times.The number of fires in the grassland in 2004 was the lowest,which was 138.From the monthly scale,the most months of fire in the Hulun Buir grassland occurred in April,May and September and October,and the number of grassland fires was the least in January and December.(2)Using the MODIS fire product data MOD14A1(Terra)and MYD14A1(Aqua)from 2000 to 2010 and related climatic data,a monthly climate index model of the fire in the inter-monthly grassland was established based on the Logistic regression method.The results showed that the monthly maximum temperature,monthly sunshine hours and monthly mean wind speed were positively correlated with the fire climate index;monthly precipitation,monthly average temperature,monthly average relative humidity,monthly minimum relative humidity,and monthly intraday precipitation The number of days greater than or equal to 5 mm,the monthly average temperature,and the fire climate index were significantly negatively correlated.The correlation between the 2011-2014 grassland fire point data and the fire climate index during the time period was used to verify the accuracy of the model and the verification result was good(Pearson correlation coefficient was 0.578).In order to better verify the accuracy of the model,this paper The ROC(receiver operating characteristic curve)was used to verify the accuracy of the model.The area under the AUC(area under curve)was 0.845,indicating that the monthly fire climate index model is suitable for analyzing the occurrence of fire in the Inner Mongolia grassland during the month of Hulun Buir.(3)Using the MODIS fire product data MOD14A1(Terra),MYD14A1(Aqua)and climatic data from 2000 to 2010,the climatic factors that have a hysteresis effect on grassland fires were selected using cross-correlation analysis methods.Monthly maximum temperature in one month,monthly maximum temperature in the first two months,monthly average temperature in the previous month,monthly precipitation in the first two months,monthly precipitation in the first three months,and month in the previous month Average wind speed,and using the Logistic regression method to establish the grassland fire climate lagged index model,the accuracy of the grassland fire climate lagged index model is 81.4%.(4)The monthly maximum temperature,monthly sunshine hours,monthly average wind speed,monthly precipitation,monthly average temperature,monthly average relative humidity,monthly minimum relative humidity,and monthly intraday after the fire climate index and Kriging interpolation were obtained by using correlation analysis methods.The correlation coefficient between the number of days with precipitation greater than or equal to 5 mm and the monthly average temperature,the monthly maximum temperature in the previous month after Kriging interpolation,and the maximum monthly temperature in the first two months,the first one The monthly average temperature of the month,the monthly precipitation in the first two months,the monthly precipitation in the first three months,and the monthly average wind speed in the previous month have obtained the revised model of the fire climate index.And use the 2011-2014 grassland fire point data and the fire climate correction index during the time period to verify the correlation,the verification result is better(Pearson correlation coefficient is 0.684),which shows that the fire climate correction index is more suitable to reflect the Hulun Buir than the fire climate index Month grassland fire.(5)By analyzing the spatial-temporal distribution pattern of the Hulun Buir grassland fire climate index and the fire climate correction index from January to December in 2011-2014,it was found that the months with higher fire climate index and fire climate correction index occurred in March,April,and May.In September and October,the fire climate index and fire climate correction index in other months were relatively low,and the area with high fire climate index and fire climate correction index were mainly distributed in the eastern and western regions of Hulun Buir.The fire climate index in the central region and the fire climate correction index is the lowest.Compared with the fire climate index,in March,the area of the high fire climate correction index area increased in the eastern and western regions of Hulun Buir;the area of extremely high fire climate correction index area in April,May,September and October,there are various degrees of increase in the eastern and western regions of Hulun Buir;the area of the extremely low-fire climate correction index in July and August has been reduced to varying degrees in the central region of Hulun Buir,and the area and height of the medium fire climate correction index region have increased.The area of the fire climate correction index area has decreased in the western region of Hulun Buir. |
PDF Full Text Request