Forest fire danger grade forecasting method based on subcompartment scale

Abstract

The invention discloses a forest fire danger grade forecasting method based on a subcompartment scale, and belongs to the technical field of forest fire prevention, the forest fire danger grade forecasting formula is P = (T + S) * M, P is a forest fire index; t is a combustible type index; s is a forest fire source index; m is the moisture content index of the fine dead combustible. The problems that existing forest fire risk prediction is low in fineness, low in precision and complex in prediction process are solved.

Description

Technical Field

[0001] The invention relates to the technical field of forest fire prevention, and in particular to a forest fire risk level forecasting method based on a small class scale. Background Art

[0002] The forest and grassland fire risk warning work has long been constrained by technical theories and technical means, and has not effectively solved the problem of refined forest fire risk prediction and warning. The forest fire risk level forecast is issued by the meteorological department. The forecast calculation method is determined according to the national standard "Forest Fire Risk Meteorological Level" (GB / T36743-2018), and the scope of application is based on the county as the minimum area for forecasting. The comprehensive forest fire risk forecast model adopted by the meteorological departments of some provinces and cities is based on.

[0003] The Forest Fire Danger Meteorological Grade introduces the consideration of combustible factors and human factors, and the results are more refined in spatial distribution and the forecast accuracy is also improved. However, there are problems with insufficient precision, which cannot be used as a small class unit for forecasting, and cannot provide guidance for grassroots management work; second, the problem of insufficient accuracy is that the granularity of combustible type zoning is relatively coarse, and the influence of the flammability of different tree species and terrain factors is not considered, and the differences in fire risk levels caused by different fire behaviors in different forest stands cannot be reflected.

[0004] The invention patent named "Method for zoning forest fire risk at small-scale for artificial forests based on combustible hazard" discloses a method for fine-scale division of forest fire risk based on the combustible hazard level of artificial forests. This invention is different from the fire risk level forecast. It only divides forest fire risk areas and does not predict fire risk. In addition, the current large-scale forest fire risk forecast is also difficult to apply to small-scale risk forecast. Conventional forest fire risk forecasting is complex and difficult to use in practice. Therefore, it is necessary to establish an accurate forest fire risk level forecasting model to reduce fire risks while providing a more simplified and convenient forecasting method. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a forest fire risk forecasting method based on small class scale, so as to solve the current problems that it is difficult to predict forest fire risk for small classes, as well as the low precision and accuracy of forest fire risk prediction and the complex prediction process.

[0006] The present invention solves the above technical problems through the following technical solutions:

[0007] A forest fire risk level forecasting method based on small class scale, the forest fire risk level forecasting first needs to calculate the forest land fire index P, the formula is: P = (T + S) × M, where P: forest land fire index; T: combustible type index; S: forest land fire source index; M: fine dead combustible moisture index, and the forest fire risk level forecast result is obtained based on the relationship between the calculated result P forest land fire index and the forest fire risk level.

[0008] Furthermore, the calculation method of the combustible type index T is as follows:

[0009] T = 0.8K + 0.2D, where K is the flammability of the forest stand; D is the terrain condition;

[0010] The calculation method of the flammability K value of the forest stand is as follows:

[0011] According to the National Forest Fire Hazard Zoning Level (LY / T1063-2008), the burning difficulty of dominant tree species (groups) in the small class area is divided into three categories: difficult to burn, flammable, and flammable; according to the small class canopy density level, it is divided into four categories: high canopy density (>0.7), medium canopy density (0.4-0.69), low canopy density (0.2-0.39), and no canopy density (<0.2);

[0012] The combustible type and K value are generated according to the above-mentioned combustion difficulty and canopy density. The specific K value is as follows:

[0013]

[0014]

[0015] The terrain condition D is calculated as follows: D=D1+D2+D3, where D1 is the slope score, D2 is the slope aspect score, and D3 is the slope position score;

[0016] The specific D1 slope value is selected as follows:

[0017]

[0018] The specific D2 slope score is selected as follows:

[0019]

[0020]

[0021] The specific D3 slope score is selected as follows:

[0022]

[0023] Further, the calculation method of the fire source index S is as follows: Among them, w1 is 0.7, w2 is 0.3, C1 is the statistical value of the number of fire points in the small class area, and C2 is the statistical value of the number of fire points in the small class and buffer area.

[0024] Furthermore, the statistics of the number of fire points in the C1 sub-class area, C2 sub-class area and buffer area are counted using the GIS geographic information system.

[0025] Furthermore, the number of fire points in the C2 small class and buffer area refers to: the total number of fire points in the small class area plus the buffer distance of 200m around the small class area.

[0026] Furthermore, the fine dead combustible moisture content index M can be obtained by using the actually measured fine dead combustible moisture content or by using the fine combustible moisture code FFMC in the Canadian Forest Fire Weather Index system FWI.

[0027] Of course, in addition to the method for calculating the fine dead combustible matter moisture content index M disclosed in the present invention, the M value calculated by other fine dead combustible matter moisture content prediction models can also be used.

[0028] Furthermore, the M value is calculated using the actual measured value to calculate the P value, and the obtained value is the forest fire risk level of the day; the M value is predicted using the FFMC method, and the obtained value is the predicted forest fire risk level.

[0029] Furthermore, the forest fire risk level is divided according to the P forest land fire index, and the specific division method is as follows:

[0030] (1) Calculate the fuel type index T and forest fire source index S of all local forest classes;

[0031] (2) According to the relationship between moisture content of combustibles and flammability (as shown in the following table), different moisture contents of dead combustibles in the table (10%, 16%, 25%, 35%) are used as the fine dead combustible moisture content index M (10, 16, 25, 35) for calculation. Then, the combustible type index T and forest fire source index S are respectively substituted into the formula P = (T + S) × M to obtain the set of local forest class forest fire index P values. Then, the natural breakpoint method is used for grouping processing to find the four group boundary values ​​of P.

[0032] The relationship between the moisture content of combustibles and their flammability is as follows:

[0033]

[0034] (3) Based on the calculated group boundary values ​​of the forest fire index P, a corresponding table of the relationship between the forest fire index P and the forest fire risk level is reconstructed.

[0035] The combustible type index T value of a certain area is usually calculated once every five years, the forest fire source index S is usually calculated once a year, and the calculated forest fire index P and forest fire risk level relationship table can also be used for one year, which greatly simplifies the forest fire risk level prediction calculation amount disclosed in the present invention and makes the prediction method more practical.

[0036] Furthermore, the prediction method is applicable to the prediction of forest fire risk level at the scale of small forestry classes.

[0037] Beneficial effects:

[0038] The present invention calculates the T value and S value of the zoning sub-class by dividing the forest combustible types and combining the historical fire observation data; collects the combustible moisture content data by establishing the combustible moisture content monitoring point, and calculates the M value according to the combustible moisture content observation result or the FFMC calculation method; and divides the ignition index P value by applying the natural breakpoint method, thus realizing the prediction of forest fire risk level. From the perspective of spatial scale, it can be applied to less than 1km×1km, and the minimum reaches 66m 2 The fire danger level forecast of the small class area. Compared with the fire weather forecast results of the same period, the fire danger level forecast results obtained in this study show a trend, and there will be no sudden change from level 1 to level 5 in the forest fire danger level, which better reflects that the change of fire danger is a process of accumulation and change of natural environmental factors. In addition, the forest fire danger level forecast method established in this study has the characteristics of clear logic, easy implementation, good versatility, strong practicality and high accuracy. The forecast results can better guide the decision-making and management of grassroots forest fire prevention work. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 : Forest fire risk level forecast released by the National Early Warning Information Release Center. DETAILED DESCRIPTION

[0040] The present invention will be described in detail below with reference to specific embodiments:

[0041] The present invention discloses a method for predicting forest fire risk level based on small class scale. The specific prediction method is as follows:

[0042] P = (T + S) × M, where P is the forest fire index; T is the combustible material type index; S is the forest fire source index; M is the moisture content index of fine dead combustible materials.

[0043] The calculation method of T value is as follows:

[0044] T=0.8K+0.2D, where K is the flammability of the forest stand and D is the terrain condition.

[0045] According to the National Forest Fire Hazard Zoning Level (LY / T1063-2008), the burning difficulty of dominant tree species in the small class area is divided into three categories: difficult to burn, flammable, and flammable; according to the canopy density level of the small class, it is divided into four categories: high canopy density (>0.7), medium canopy density (0.4-0.69), low canopy density (0.2-0.39), and no canopy density (<0.2); according to the above-mentioned burning difficulty and canopy density, the combustible type and K value are generated. The specific K value is shown in Table 1:

[0046] Table 1 Combustible types and K values

[0047]

[0048]

[0049] The calculation method of terrain condition D is as follows:

[0050] D=D1+D2+D3, where D1 is the slope score as shown in Table 2, D2 is the slope direction score as shown in Table 3, and D3 is the slope position score as shown in Table 4.

[0051] Table 2D1 Slope score

[0052]

[0053] Table 3D2 Slope Aspect Score

[0054]

[0055]

[0056] Table 4D3 Slope Points

[0057]

[0058] The calculation method of the fire source index S is as follows:

[0059] Among them, w1 is 0.7, w2 is 0.3, C1 is the statistical value of the number of fire points in the small class area, and C2 is the statistical value of the number of fire points in the small class and buffer area.

[0060] Specifically, C1 is the statistical value of the number of fire points in the small class area and C2 is the statistical value of the number of fire points in the small class and buffer area, which are counted using the geographic information system QIS software. When calculating the data, the data for the whole year can be preferably used for statistics.

[0061] Statistics of the number of fire points in the small class area (C1): Import the fire data into GIS in the form of point features to form a point layer, and spatially overlay it with the polygon layer of the small class area. By applying the built-in "Count the number of points in the polygon" tool in GIS, the number of points in the point layer in the polygon layer is calculated, and then the C1 value of each small class is counted.

[0062] Number of fire points in small classes and buffer areas (C2): First, create a buffer zone of a certain distance for each small class. According to the experience of fire source control in urban mountainous areas, the buffer distance is set to 200m. Then, the C1 method is used to count the total number of fires in small classes and buffer areas to obtain the C2 value.

[0063] The calculation method of moisture content index M of fine dead combustibles is as follows:

[0064] 1. Use the combustible real-time monitoring sensor to measure the moisture content value at 14 o'clock × 100 to get the M value;

[0065] 2. Use the method of fine combustible moisture code FFMC in the Canadian Forest Fire Weather Index System FWI to calculate the M value. For specific calculation methods, refer to "Development and Structure of the Canadian Forest Fire Weather Index System CEVan Wagner Canadian Forestry Service Forestry Technical Report 35. Ottawa 1987" or "Forest Fire Ecology and Management (Revised Edition) Edited by Hu Haiqing, China Forestry Publishing House, 2005.8."

[0066] Substituting the M value measured by method 1 into the formula to calculate the P value gives the forest fire risk level for the day;

[0067] The M value predicted by the FFMC method of Method 2 is the predicted forest fire danger level.

[0068] The forest risk level is divided according to the P forest land fire index. The specific division method is as follows:

[0069] (1) Calculate the fuel type index T and forest fire source index S of the local forest class;

[0070] (2) As shown in Table 5, based on the relationship between the moisture content of combustibles and flammability, the moisture content of combustibles is regarded as the moisture content index M of fine dead combustibles. The combustible type index T, forest fire source index S value, and the moisture content index M of fine dead combustibles of each small class are respectively substituted into the formula P = (T + S) × M to obtain the set of forest fire index P. Here, the moisture content index of fine dead combustibles M = 10 (combustible moisture content is 10%), 16 (combustible moisture content is 16%), 25 (combustible moisture content is 25%), and 35 (combustible moisture content is 35%) are used to calculate the P value. The natural breakpoint method is used to find the four grouping boundary values ​​of P.

[0071] Table 5 Relationship between moisture content M of combustibles and flammability

[0072]

[0073] (3) Reconstruct the correspondence table between the forest fire index P and the forest fire risk level according to the group boundary value of P, and find the corresponding group according to the actual calculated P value to obtain the forest fire risk level forecast result.

[0074] Embodiment 1:

[0075] 1. Region: A small class of Jinpenhu Management Station, Chashan Zhuhai Forest Farm, Yongchuan District, with an area of ​​about 20 mu, dominant tree species: Masson pine, canopy density 0.7, calculate the forest fire risk level on August 15, 2023. Check the meteorological records of Yongchuan District at 14:00 on August 15, 2023, the temperature was 38 degrees, the air humidity was 13%, and there was no precipitation. The moisture content sensor of the combustible material in the forest measured the moisture content of the fine dead combustible material on the surface at 14:00 on that day.

[0076] (1) According to the type of combustible materials, it belongs to the high density flammable type, K value = 0.72; the terrain conditions are slope, south slope, and downhill slope, that is, D1 = 0.6, D2 = 0.86, and D3 = 0.8.

[0077] (2) Based on the data for the whole year of 2022, the statistical value of the number of fire points in the small class area is C1=0, and the statistical value of the number of fire points in the small class and buffer area is C2=1.

[0078] (3) Using the actual measured value of 10% of the moisture content sensor in the combustible material, the M value on August 15, 2023 is obtained.

[0079] =10.

[0080] 2. Substitute the above values ​​into the following formula:

[0081] (1) Formula 1: T = 0.8K + 0.2D;

[0082] (2) Formula 2: Where w1 is 0.7 and w2 is 0.3;

[0083] The obtained T value = 1.028, S value = 0.06069784, and then substitute the above results into Formula 3: P = (T+S)×M = (1.028+0.06069784)×10, and the calculated P value on August 15, 2023 is 10.8870.

[0084] 3. The corresponding calculation of the P forest land fire index of all forest land classes in Yongchuan area (here refers to the experimental demonstration area) and the forest fire risk level table is as follows:

[0085] (1) Substitute the combustible type index T and fire source index S of each class into formula 3;

[0086] (2) Refer to Table 5 and substitute the M values ​​of 10, 16, 25, and 35 into Formula 3 in sequence;

[0087] (3) According to Formula 3: P = (T + S) × M, the P value set of all small classes is obtained.

[0088] Taking a small class at the Jinpenhu Management Station of the Chashan Zhuhai Forest Farm in Yongchuan District as an example, the calculation is as follows:

[0089] When M=10, P=(1.028+0.06069784)×10=10.8869784;

[0090] When M=16, P=(1.028+0.06069784)×16=17.41916544;

[0091] When M=25, P=(1.028+0.06069784)×25=27.217446;

[0092] When M=35, P=(1.028+0.06069784)×35=35=38.1044244.

[0093] There are a total of 7635 small classes in Yongchuan area (experimental demonstration area), and the corresponding P values ​​are calculated for all of them. Then, after being processed by the natural breakpoint method, the relationship table of combustible moisture content-flammability performance-forest fire index P value-forest fire risk level in Yongchuan area (experimental demonstration area) is obtained, as shown in Table 6:

[0094] Table 6 Correspondence between forest fire index P and forest fire risk level in Yongchuan area

[0095]

[0096] 4. The P value calculated in step 2 on August 15, 2023 = 10.8870 is searched with the P value in Table 6, and the results are as follows: The forest fire risk level of a small class in the Jinpenhu Management Station of Chashan Zhuhai Forest Farm in Yongchuan District on August 15, 2023 is level 5, which is consistent with the fire risk weather level on August 15, 2023 ( Figure 1 ) can be compared to know that the overall forest fire hazard meteorological level of Yongchuan District released by the National Early Warning Information Release Center is a higher forest fire hazard meteorological level, which belongs to the forest fire hazard meteorological level forecast for a large area. However, for a small class of Jinpenhu Management Station of Chashanzhuhai Forest Farm in Yongchuan District, the forecast is inaccurate. According to the results calculated by the present invention, it should belong to the fifth level of forest fire hazard risk.

[0097] Embodiment 2:

[0098] 1. Region: A small class at Jinpenhu Management Station, Chashan Zhuhai Forest Farm, Yongchuan District, with an area of ​​about 20 mu, dominant tree species: Pinus massoniana, canopy density 0.7, calculate the forest fire risk level on August 16, 2023. Check the weather forecast for Yongchuan District on August 16, 2023, the highest temperature is 37 degrees, the air humidity is 20%, and there is no precipitation.

[0099] (1) According to the type of combustible materials, it belongs to the high density flammable type, K value = 0.72; the terrain conditions are slope, south slope, and downhill slope, that is, D1 = 0.6, D2 = 0.86, and D3 = 0.8.

[0100] (2) Based on the data for the whole year of 2022, the statistical value of the number of fire points in the small class area is C1=0, and the statistical value of the number of fire points in the small class and buffer area is C2=1.

[0101] (3) The fine combustible moisture content index was calculated using the method of fine combustible moisture code FFMC in the Canadian Forest Fire Weather Index System FWI recorded in the reference “Development and Structure of the Canadian Forest Fire Weather Index System CEVan Wagner Canadian Forestry Service Forestry Technical Report 35. Ottawa 1987”, and the M value on August 16, 2023 was 7.198834186.

[0102] 2. Substitute the above values ​​into the following formula:

[0103] (1) Formula 1: T = 0.8K + 0.2D;

[0104] (2) Formula 2:

[0105] The obtained T value = 1.028, S value = 0.06069784, and then substitute the above results into Formula 3: P = (T+S)×M, and the calculated P value on August 16, 2023 = 7.8374.

[0106] 3. The combustible material type index T value in Yongchuan area is usually calculated once every five years, and the forest fire source index S is usually calculated once a year. The calculated forest fire index P and forest fire risk level relationship table can also be used for one year. Therefore, the P value = 7.8374 calculated in step 2 on August 16, 2023 is searched with the P value in Table 6 on the correspondence between the forest fire index P and forest fire risk level in Yongchuan area. The results are as follows: It is predicted that the forest fire risk level of a small class in the Jinpenhu Management Station of Chashan Zhuhai Forest Farm, Yongchuan District, on August 16, 2023 is level five.

[0107] Fire Danger Weather Level on August 16, 2023 ( Figure 1 ) It can be seen from the comparison that the overall forest fire risk meteorological level of Yongchuan District released by the National Early Warning Information Release Center is a high forest fire risk meteorological level, which belongs to the forest fire risk meteorological level forecast for a large area, but it is not accurate for a small class of Jinpenhu Management Station of Chashan Zhuhai Forest Farm in Yongchuan District. According to the prediction results of the present invention, it should belong to the fifth level of forest fire risk. At the same time, since the P value is smaller than that on the 15th, it means that the fire risk of the small class is increasing.

[0108] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions of the present invention, which should be included in the scope of the claims of the present invention. The techniques, shapes, and structural parts not described in detail in the present invention are all known technologies.

Claims

1. A method for predicting forest fire risk level based on small class scale, characterized in that: The forest fire risk level prediction first requires the calculation of the forest fire index P, the formula is: P = (T + S) × M, where P: forest fire index; T: combustible type index; S: forest fire source index; M: fine dead combustible moisture index, and the forest fire risk level forecast result is obtained based on the relationship between the calculated result P forest fire index and the forest fire risk level.

2. A method for forecasting forest fire risk level based on small class scale according to claim 1, characterized in that: The calculation method of the combustible type index T is as follows: T = 0.8K + 0.2D, where K is the flammability of the forest stand; D is the terrain condition; The calculation method of the flammability K value of the forest stand is as follows: According to the National Forest Fire Hazard Zoning Level (LY / T1063-2008), the burning difficulty of small-scale dominant tree species (groups) is divided into three categories: difficult to burn, flammable, and flammable; according to the small-scale canopy density level, it is divided into four categories: high canopy density (>0.7), medium canopy density (0.4-0.69), low canopy density (0.2-0.39), and no canopy density (<0.2); Based on the burning difficulty of the dominant tree species (groups) in the above-mentioned small class and the canopy density of the small class forest, the fuel type and K value of the small class are generated. The specific K value is as follows: The terrain condition D is calculated as follows: D=D1+D2+D3, where D1 is the slope score, D2 is the slope aspect score, and D3 is the slope position score; The specific D1 slope value is selected as follows: The specific D2 slope score is selected as follows: The specific D3 slope score is selected as follows:

3. A method for predicting forest fire risk level based on small class scale according to claim 1, characterized in that: The calculation method of the fire source index S is as follows: Among them, w1 is 0.7, w2 is 0.3, C1 is the statistical value of the number of fire points in the small class area, and C2 is the statistical value of the number of fire points in the small class and buffer area.

4. A method for predicting forest fire risk level based on small class scale according to claim 3, characterized in that: The statistics of the number of fire points in the C1 sub-class area, C2 sub-class area and buffer area are calculated using GIS geographic information system software.

5. A method for predicting forest fire risk level based on small class scale according to claim 4, characterized in that: The number of fire points in the C2 small class and buffer area refers to the total number of fire points in the small class area plus the buffer distance of 200m around the small class area.

6. A method for forecasting forest fire risk level based on small class scale according to claim 1, characterized in that: The fine dead combustible moisture content index M can be obtained by selecting the actually measured fine dead combustible moisture content as an index or by using the fine combustible moisture code FFMC in the Canadian Forest Fire Weather Index system FWI to calculate the index.

7. A method for forecasting forest fire risk level based on small class scale according to claim 6, characterized in that: The M value is calculated by using the actual measured value to calculate the P value, and the obtained value is the forest fire risk level of the day; the M value is predicted by using the FFMC method, and the obtained value is the predicted forest fire risk level.

8. A method for forecasting forest fire risk level based on small class scale according to any one of claims 1 to 7, characterized in that: The forest fire risk level is divided according to the P forest land fire index, and the specific division method is as follows: (1) Calculate the fuel type index T and forest fire source index S of all local forest classes; (2) According to the relationship between the moisture content of combustibles and flammability, the moisture content of different combustibles in the table is used as the moisture content index M of fine dead combustibles for calculation. Then, the combustible type index T and the forest fire source index S are respectively substituted into the formula P = (T + S) × M to obtain the set of P values ​​of the local forest class. The natural breakpoint method is used for grouping processing to find the group boundary value of P. The relationship between the moisture content of combustibles and their flammability is as follows: (3) Based on the calculated group boundary values ​​of the forest fire index P, the corresponding table between the forest fire index P and the forest fire risk level is reconstructed.

9. A method for forecasting forest fire risk level based on small class scale according to any one of claims 1 to 8, characterized in that: The prediction method is suitable for predicting forest fire risk levels at the scale of small forestry classes.

Images