CN103886386A - Method for predicting manual fire day occurrence probability based on space grid - Google Patents
Method for predicting manual fire day occurrence probability based on space grid Download PDFInfo
- Publication number
- CN103886386A CN103886386A CN201410057162.9A CN201410057162A CN103886386A CN 103886386 A CN103886386 A CN 103886386A CN 201410057162 A CN201410057162 A CN 201410057162A CN 103886386 A CN103886386 A CN 103886386A
- Authority
- CN
- China
- Prior art keywords
- fire
- data
- day
- forest fire
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for predicting manual fire day occurrence probability based on a space grid, comprising steps of establishing data files and using each grid point and all fire points in a detection area as sampling points, performing multiple colinearity diagnosis through the data, choosing factors having obvious impact on the occurrence of fire through analysis of Logistic Forward Wald, establishing a fire day occurrence probability model through a Logistic model so as to analyze the manual fire day occurrence probability, determining manual fire ignition threshold through a secondary determination theory, calculating precision and performing precision analysis, and substituting day value data of important factors into the model to calculate based on GIS. As a result, seven processes of predicting manual fire day occurrence probability are finished. The invention can be used for predicting daily manual fire occurrence and can be used for predicting and evaluating the fatalness of manual fire occurrence in the future based on the weather scene data.
Description
Technical field
Patent of the present invention relates to a kind of method of predicting man caused forest fire day probability of happening, relates in particular to a kind of method based on space lattice prediction man caused forest fire day probability of happening.
Background technology
Forest fire is that world today's generation is wide, sudden strong, destructive large, disposes the comparatively difficult disaster of putting out a fire to save life and property, and along with global warming, fire has the trend of rising.China is especially big also in rising trend with fatal forest fire generation in recent years, and China's forest fire protection situation is always in severe state.Forest fire causes heavy economic losses and environmental disruption, the serious threat resident security of the lives and property.
Mainly adopt fire forecast about the means of causing danger property of forest fire forecast at present, forest fire weather forecasting and forest fire danger class forecast the only Hazard rank of prediction particular weather condition initiation forest fire.Fire forecast is mainly, according to meteorological element and vegetation, the reaction of environmental baseline is estimated to the possibility that forest fires occur, and namely whether weather condition is conducive to forest fires generation.Therefore, the pre-retribution of fire selects to reflect that weather do wet meteorological element, and the overwhelming majority is changed factor, as air themperature, relative humidity, precipitation, wind, company's number dry days etc.; Vegetation condition intends choosing the water percentage of combustible water percentage, particularly fine fuels.According to these fire danger factors, utilize mathematics or statistical method to calculate danger index, divide fire size class.
Forecast occurs in forest fire is to carry out on the basis of forecasting in fire, and forest fire forecast occurs and requires to forecast a certain the area interior forest fires of section occur sometime probability or number of times.There is forecast and consider danger that the dry wet degree variation of Changes in weather, combustible and fuel type and burning things which may cause a fire disaster occur etc. and carry out the possibility that prediction fire occurs in forest fire.Therefore, forest fire generation predictor is chosen and need to be reflected fire predictor.
Belong to fire generation forecast category for forecast occurs fire, the research of this respect still belongs to the Primary Study stage both at home and abroad at present, owing to relating to the diversity of the factor, correlativity between the factor, being difficult to of partial factors is acquired, and the inaccuracy of data etc. all has influence on the structure of forest fire man caused forest fire day occurrence Probability Model.Forecast occurs man caused forest fire is the importance that forecast occurs Forest Fire, due to complicacy and the randomness of human activity, makes to there is no at present ripe man caused forest fire forecast model products.
Summary of the invention
Adopt following technical scheme in order to solve position that man caused forest fire occurs and problem the present invention of man caused forest fire probability of happening forecast thereof.The method comprises the steps:
Step 1, definite factor and collection related data that affects forest fire, wherein test zone is divided into uniform grid, each grid cell is as a space independently, and the man caused forest fire probability of happening calculating is the man caused forest fire probability of happening of each grid cell;
Step 2, set up data file, in test zone, each lattice point and all fire point are as sampled point;
Step 3, carry out multicollinearity diagnosis by data;
Step 4, by Logistic Forward Wald analyze, choose the factor that fire is had to appreciable impact;
Step 5, set up man caused forest fire day occurrence Probability Model by Logistic model, thereby analyze man caused forest fire day probability of happening,
z=b
0+b
1x
1+b
2x
2+…+b
px
p
B
pfor coefficient or constant term, e is natural number.
Step 6, differentiate the theoretical man caused forest fire threshold values that catches fire of determining by secondary, computational accuracy, thus carry out precision analysis;
Mean value and standard deviation by catch fire sample and missing of ignition sample probability are calculated, and can obtain criterion:
In formula
and S
1, S
2be respectively mean value and the standard deviation of catch fire sample and missing of ignition sample probability.
If P (fire)>=Y
0, be judged to and catch fire;
If P (fire) is <Y
0, be judged to and do not catch fire.
Step 7, based on GIS, a day Value Data substitution model for key factor is calculated, thereby complete the probability of happening forecast of man caused forest fire day.
In above-mentioned steps two, generate man caused forest fire point location map according to the latitude and longitude coordinates of fire data.The date and the position that occur according to man caused forest fire, obtaining the each component factor of each fire FWI system and weather data processes man caused forest fire, man caused forest fire data to every day are processed, man caused forest fire point and grid are superposeed, if this fire point is positioned at certain grid, this grid differentiation is 1 (catching fire), is 0 (not catching fire) otherwise differentiate.
Further, the factor that affects forest fire in described step 1 comprises that the processing of man caused forest fire historical data, space lattice division, space length and height above sea level calculating, vegetation data processing, grid longitude and latitude, process meteorological data, the each component factor of FWI are calculated, the Julian date.
Above factor specific explanations is:
(1) man caused forest fire historical data processing
In statistical forecast region man caused forest fire day frequency and date of generation, and man caused forest fire generation longitude and latitude.
(2) space lattice is divided
Estimation range is divided into uniform grid, and each grid cell is as a space independently, and the man caused forest fire probability of happening calculating is the man caused forest fire probability of happening of each grid cell.
(3) space length and height above sea level calculate
Carry out distance with man caused forest fire point position data and highway, railway and residential block respectively and calculate, calculate the effectively bee-line of fire point position, and man caused forest fire is put to position data and altitude figures stack, generate the data of bee-line and sea level elevation.
(4) vegetation data processing
Man caused forest fire is put to position data and combustible substance distribution figure superposes, carry out intersection operation, obtain fuel type of each fire point position.
The space lattice data of generation and vegetation data are superposeed, calculate the vegetation pattern of each grid inner area maximum as the vegetation pattern of this grid, generate vegetation data Layer, respectively every kind of vegetation pattern is set to a code.
(5) grid longitude and latitude
Different longitudes and latitude produce different impacts to weather, and have affected to a certain extent the distribution of forest.Longitude and latitude, has determined the locus in a certain region, local Climatic and vegetation is distributed and produces long-term impact simultaneously, the latitude and longitude coordinates using the latitude and longitude coordinates at each cell center as cell.
(6) process meteorological data
Continuous meteorological day Value Data, comprises temperature, precipitation, phase water humidity, wind speed.
(7) the each component factor of FWI is calculated
In model, need to input the output parameter in forest fires weather index FWI system according to result, be respectively: fine fuels humidity codes FFMC, duff humidity codes DMC, arid code DC, initial rate of propagation ISI, combustible index of bunching BUI, fiery weather index FWI.According to carrying out Collinearity Diagnosis Analysis result with correlation factor, the each component factor of FWI is selected.
FFMC, DMC and DC represent that different classes of Moisture of Forest Flammable Matter changes, and ISI, BUI and FWI are fire behavior indexs.
(8) the Jul ian date
Reflect the impact of seasonal variation on man caused forest fire probability of happening.
Further, in described step 2, the index of each fire point and grid is set up in the definite employing whether each grid is caught fire, if sample on the same day, this grid state of catching fire is made as 1, otherwise is made as 0, final table data file, the interpolated data of a certain day set up.
Further, in described step 2, the data of the collection to sampled point are vegetation information, altitude information, from mankind's accumulation area minimum distance information etc.
Further, in described step 3, carry out in multicollinearity diagnosis the factor of influence of selecting expansion factor to be less than 10 by data.
Because the collinearity of independent variable can reduce the precision of model, in order to improve the precision of modeling, reduce the collinearity between different variablees in model, before modeling, to carry out multicollinearity check to selected variable for this reason.According to the variance inflation factor VIF of variable, collinearity between the larger explanation related coefficient of VIF value is larger, VIF value is less than 10 under normal circumstances, representing does not have obvious correlativity between each variable, tolerance is less, multicollinearity is more serious, and in addition also can be by eigenvalue of maximum time, the variation proportion of each variable carrys out the correlativity between judgment variable.It is generally acknowledged that VIF<10 thinks that linear dependence is not obvious, the impact that these variablees occur fire is different.
Beneficial effect of the present invention is:
The method has been considered to affect the key factor that forest fire occurs, and the probability that has solved the man caused forest fire every day generation of a certain position, a certain region is how many problem, is the core of forest fire generation forecast system.The method has overcome traditional fire size class and has lacked the shortcoming that quantification is described, and can judge the possibility that man caused forest fire danger occurs in certain area intuitively, and it is more accurate to have, feature more directly perceived.Both can forecast for daily man caused forest fire probability of happening, also can be for following causing danger property of man caused forest fire being assessed based on Climate Scenarios data.The structure of man caused forest fire generation model need be based on specific region source data, and build model and be applicable to the specific region corresponding with source data, along with the prolongation of time, the increase of forest fire historgraphic data recording, the precision of model will further improve.
Accompanying drawing explanation
Fig. 1 is the inventive method step schematic diagram;
Fig. 2 is man caused forest fire day of the present invention probability of happening simulation drawing.
Embodiment
Further illustrate technical scheme of the present invention below in conjunction with accompanying drawing and by embodiment.
In conjunction with Fig. 1 and Fig. 2, (comprise the large portion of the Greater Hinggan Mountains in Heilongjiang and Daxinganling, Inner Mongolia) take Daxing'an Mountainrange and carry out the probability of happening modeling of man caused forest fire day and simulation as example.
Step 1 101, definite factor and collection related data that affects forest fire, wherein test zone is divided into uniform grid, each grid cell is as a space independently, and the man caused forest fire probability of happening calculating is the man caused forest fire probability of happening of each grid cell.
(1) by estimation range take 20km as elementary cell is divided into uniform grid, each grid cell is as a space independently.
(2) the 1972-2006 man caused forest fire data of collection Heilongjiang Province and Daxinganling District, Inner Mongolia Autonomous Region, comprise date, fire number of times, man caused forest fire point longitude and latitude data that man caused forest fire occurs.Generate man caused forest fire point location map according to the latitude and longitude coordinates of fire data.
(3) supposing under the constant condition in combustible substance distribution, highway, railway and residential block, respectively with effectively calculate fire point data and DEM, combustible substance distribution, highway, railway and residential block, calculate effectively fuel type and the bee-line of fire point position, generate the field that comprises fuel type, bee-line and sea level elevation.
(4) in order to reduce operand, the each component factor of meteorological factor and FWI system adopts inverse distance weighting (Inverse Distance Weighting, IDW) interpolation method, IDW interpolation method is determinacy interpolation method, suppose observed reading and distance dependent at a distance, distance is far away, contributes less.
K
jvalue be a regulator, to guarantee that weight sum is as 1.If a=1,
(5) collect the weather data of time period identical with man caused forest fire data, weather data employing mean value calculates.Weather data is used 1972-2006 country Value Data of basic website day, has 10 websites, in order to improve the effect of interpolation, has also selected 3 websites at survey region periphery.Main field comprises temperature on average, precipitation, relative humidity, mean wind speed.The each component factor of FWI is calculated.
(6) computing formula of the each component factor of FWI system is with reference to Equations and FORTRAN program for the Canadian Forest Fire Weather Index System (Canadian forest fires weather index system equation and FORTRAN routine) and two books of Development and structure of the Canadian forest fire weather index system (system development of Canadian forest fires weather index and structure).Also can adopt Canadian Prometheus fire spread software to carry out secondary development calculating based on COM (the Component Object Model) technology.
(7) if the 1972-2006 data of all days are processed, data volume is too large, therefore, only chooses the date of breaking out of fire, weather data to this day and the FWI system index of correlation are carried out interpolation processing, and its cell size is consistent with position and foundation drawing.The each component factor of meteorological factor to all dates of catching fire of 1972-2007 and FWI system is carried out interpolation, and the date that wherein man caused forest fire occurs has 873, and there are 10 variablees on each date, carries out interpolation 8730 times.
(8) grid data and 1:100 ten thousand vegetation data are superposeed, calculate the vegetation pattern of each grid inner area maximum as the vegetation pattern of this grid, generate vegetation data Layer, respectively every kind of vegetation pattern is set to a code.Constant in the situation that, fire is put to position data with combustible substance distribution figure superposes at supposition fuel type, carry out intersection operation, obtain each fire and put fuel type of position.The height above sea level of each fire point position is read in fire point position distribution data, generates the height above sea level of each cell.
(9) different longitudes and latitude produce different impacts to weather, and have affected to a certain extent the distribution of forest.Longitude and latitude, has determined the locus in a certain region, local Climatic and vegetation is distributed and produces long-term impact simultaneously, the latitude and longitude coordinates with the latitude and longitude coordinates at each cell center as cell.
Step 2 102, set up data file, in test zone, each lattice point and all fire point are as sampled point.
In survey region, each lattice point is as a sampled point, all fire points are also as sampled point simultaneously, set up from the day data file of 1972-2006, data file comprises the vegetation information, altitude information of sampling point, from mankind's accumulation area minimum distance information etc.The index of each fire point and grid is set up in definite employing of whether catching fire for each grid, if index value is identical, is sample on the same day simultaneously, and this grid state of catching fire is made as 1, otherwise is made as 0, final table data file, the interpolated data of a certain day set up.
Step 3 to five 103-105, carry out multicollinearity diagnosis by data; Analyze by Logistic Forward Wald, choose the factor that fire is had to appreciable impact; Set up man caused forest fire day occurrence Probability Model by Logistic model, thereby analyze man caused forest fire day occurrence Probability Model structure.
According to the result of collinearity check, have 7 factor pair forest fire probability of happening and exert an influence, they are temperature on average, daily precipitation amount, relative humidity, mean wind speed, FFMC, DC and FWI.
Adopt Logi stic Forward Wald method to analyze correlated variables, remove the fire inapparent variable that makes a difference, filter out fiery probability of happening is affected to significant variable, build Logistic equation, each regression coefficient and inspection parameter are as table 1.
Table 1 regression coefficient and inspection parameter
Z=-0.0111RHM+0.0825Rain+0.0496FFMC-0.00241S1-0.00235Elev-0.000022Dist+0.126Long-0.5i511Lat+2.256
In formula, Rain is daily precipitation amount, mm; RHM is relative humidity, %; Elev is height above sea level, m; Dist is from mankind's accumulation area minimum distance, m; Long is longitude, °; Lat is latitude, °.
Step 6 106, differentiate the theoretical man caused forest fire threshold values that catches fire of determining by secondary, computational accuracy, thus carry out precision analysis precision analysis
Table 2 basic statistics amount
Respectively the sample catching fire and non-ignitable sample are added up, calculated basic statistics amount (table), according to the secondary method of discrimination whether catching fire, calculating judgment threshold is 0.0019.Man caused forest fire model again substitution is analyzed to data, fiery probability of happening is calculated, the threshold value occurring according to fire is added up catch fire sample and the sample that do not catch fire respectively, calculates the accuracy rate of judgement as table.
Table 3 accuracy rate statistics
Can find out that in the sample that catches fire, being judged as the sample size catching fire is 785, accuracy rate is 60.02%, and it is 821062 that the sample that do not catch fire is judged as the sample size not catching fire, and accuracy rate is 79.54%.
Step 7 107, based on GIS, a day Value Data substitution model for key factor is calculated, thereby judge.Choosing 1982-6-15 is example, and man caused forest fire occurrence Probability Model is simulated, and can evaluate each grid man caused forest fire probability of happening in region.
The above; be only preferably embodiment of the present invention, but protection scope of the present invention is not limited to this, any people who is familiar with this technology is in the disclosed technical scope of the present invention; the variation that can expect easily or replacement, within all should being encompassed in protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.
Claims (5)
1. the method based on space lattice prediction man caused forest fire day probability of happening, is characterized in that, should
Method comprises the steps:
Step 1, definite factor and collection related data that affects forest fire, wherein test zone is divided into uniform grid, each grid cell is as a space independently, and the man caused forest fire probability of happening calculating is the man caused forest fire probability of happening of each grid cell;
Step 2, set up data file, in test zone, each lattice point and all fire point are as sampled point;
Step 3, carry out multicollinearity diagnosis by data;
Step 4, by Logistic Forward Wald analyze, choose the factor that fire is had to appreciable impact;
Step 5, set up man caused forest fire day probability of happening Logistic model
z=b
0+b
1x
1+b
2x
2+…+b
px
p
B
pfor coefficient or constant term, e is natural number;
Step 6, differentiate the theoretical man caused forest fire threshold values that catches fire of determining by secondary,
Mean value and standard deviation by catch fire sample and missing of ignition sample probability are calculated, and can obtain criterion:
In formula
and S
1, S
2be respectively mean value and the standard deviation of catch fire sample and missing of ignition sample probability.
If P (fire)>=Y
0, be judged to and catch fire;
If P (fire) is <Y
0, be judged to and do not catch fire;
Step 7, based on GIS, a day Value Data substitution model for key factor is calculated, thereby complete the probability of happening forecast of man caused forest fire day.
2. a kind of method based on space lattice prediction man caused forest fire day probability of happening as claimed in claim 1, it is characterized in that, the factor that affects forest fire in described step 1 comprises that the processing of man caused forest fire historical data, space lattice division, space length and height above sea level calculating, vegetation data processing, grid longitude and latitude, process meteorological data, the each component factor of FWI are calculated, the Julian date.
3. a kind of method based on space lattice prediction man caused forest fire day probability of happening as claimed in claim 1, it is characterized in that, in described step 2, the index of each fire point and grid is set up in the definite employing whether each grid is caught fire, if sample on the same day, this grid state of catching fire is made as 1, otherwise is made as 0, final table data file, the interpolated data of a certain day set up.
4. a kind of method based on space lattice prediction man caused forest fire day probability of happening as claimed in claim 1, is characterized in that, in described step 2, the data of the collection to sampled point are vegetation information, altitude information, from mankind's accumulation area minimum distance information etc.
5. a kind of method based on space lattice prediction man caused forest fire day probability of happening as claimed in claim 1, is characterized in that, in described step 3, carries out in multicollinearity diagnosis the factor of influence of selecting expansion factor to be less than 10 by data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410057162.9A CN103886386A (en) | 2014-02-20 | 2014-02-20 | Method for predicting manual fire day occurrence probability based on space grid |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410057162.9A CN103886386A (en) | 2014-02-20 | 2014-02-20 | Method for predicting manual fire day occurrence probability based on space grid |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103886386A true CN103886386A (en) | 2014-06-25 |
Family
ID=50955268
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410057162.9A Pending CN103886386A (en) | 2014-02-20 | 2014-02-20 | Method for predicting manual fire day occurrence probability based on space grid |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103886386A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104463883A (en) * | 2014-12-17 | 2015-03-25 | 国家电网公司 | Method for evaluating forest fire spreading risk of power transmission channel |
| CN108765836A (en) * | 2018-05-22 | 2018-11-06 | 深圳源广安智能科技有限公司 | A kind of forest fire early-warning system based on wireless sensor network |
| CN113553754A (en) * | 2020-04-23 | 2021-10-26 | 中国石油化工股份有限公司 | Memory, fire risk prediction model construction method, system and device |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101620451A (en) * | 2008-06-30 | 2010-01-06 | 中国林业科学研究院森林生态环境与保护研究所 | Quick search ruler for forest fire danger classes |
-
2014
- 2014-02-20 CN CN201410057162.9A patent/CN103886386A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101620451A (en) * | 2008-06-30 | 2010-01-06 | 中国林业科学研究院森林生态环境与保护研究所 | Quick search ruler for forest fire danger classes |
Non-Patent Citations (1)
| Title |
|---|
| 王明玉: "气候变化背景下中国林火响应特征及趋势", 《中国博士学位论文全文数据,农业科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104463883A (en) * | 2014-12-17 | 2015-03-25 | 国家电网公司 | Method for evaluating forest fire spreading risk of power transmission channel |
| CN104463883B (en) * | 2014-12-17 | 2017-05-10 | 国家电网公司 | Method for evaluating forest fire spreading risk of power transmission channel |
| CN108765836A (en) * | 2018-05-22 | 2018-11-06 | 深圳源广安智能科技有限公司 | A kind of forest fire early-warning system based on wireless sensor network |
| CN113553754A (en) * | 2020-04-23 | 2021-10-26 | 中国石油化工股份有限公司 | Memory, fire risk prediction model construction method, system and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Understanding the spatiotemporal links between meteorological and hydrological droughts from a three‐dimensional perspective | |
| Davidson et al. | Annual prediction of shoreline erosion and subsequent recovery | |
| Beguería et al. | Assessing trends in extreme precipitation events intensity and magnitude using non-stationary peaks-over-threshold analysis: a case study in northeast Spain from 1930 to 2006 | |
| CN103279671B (en) | Municipal water disaster Risk Forecast Method based on RBF neural-cloud model | |
| Brönnimann et al. | Extreme winds at northern mid-latitudes since 1871 | |
| Huang et al. | An analytical comparison of four approaches to modelling the daily variability of solar irradiance using meteorological records | |
| Bao et al. | Coupling ensemble weather predictions based on TIGGE database with Grid-Xinanjiang model for flood forecast | |
| CN102999694B (en) | A kind of mountain region disaster takes place frequently district's Risk Evaluation of Debris Flow method | |
| Nester et al. | Flood forecast errors and ensemble spread—A case study | |
| Brocca et al. | Application of a model-based rainfall-runoff database as efficient tool for flood risk management | |
| CN103886385A (en) | Method for predicting forest fire hazard day occurrence probability | |
| CN106845080B (en) | Based on the modified Scene Tourist meteorological disaster intelligent Forecasting of difference | |
| CN113779760B (en) | Power-statistics combined season climate prediction method based on predictable climate mode | |
| Tonn et al. | Hurricane Isaac: a longitudinal analysis of storm characteristics and power outage risk | |
| Zin et al. | Statistical distributions of extreme dry spell in Peninsular Malaysia | |
| Chen et al. | Variability of seasonal precipitation extremes over China and their associations with large‐scale ocean–atmosphere oscillations | |
| CN117009735A (en) | High-strength forest fire occurrence probability calculation method combining BiLSTM and nuclear density estimation | |
| Guan et al. | Development of verification methodology for extreme weather forecasts | |
| Liu et al. | The joint return period analysis of natural disasters based on monitoring and statistical modeling of multidimensional hazard factors | |
| CN103886177A (en) | Lightning fire daily occurrence probability predicting method based on space grids | |
| CN103257000B (en) | Temperature extreme-value prediction method for bridge structure sunshine effect analysis | |
| Samiran Das et al. | Assessment of uncertainty in flood flows under climate change impacts in the Upper Thames River basin, Canada. | |
| CN103886386A (en) | Method for predicting manual fire day occurrence probability based on space grid | |
| CN103353295A (en) | Method for accurately predicating vertical deformation of dam body | |
| Ujjwal et al. | A probability-based risk metric for operational wildfire risk management |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140625 |