30 results about "Rainfall index" patented technology

The invention discloses a prediction method for a power transmission line mountain fire caused by multiple factors. The prediction method comprises the steps that (1) a grid is divided into prediction regions; (2) historical fire point density A is calculated; (3) current fire point density B is calculated; (4) the rainfall index C in every day of the following three to seven days is calculated; (5) the historical rainfall index D is calculated; (6) the high-frequency period E of industrial and agricultural fires is found out; (7) the ground surface vegetation index F is calculated; (8) mountain fire prediction density X is calculated. By means of the prediction method, mountain fire occurrences in the following three to seven days can be predicted; the operability is high; the accuracy rate of forecasting is high, quantitative prediction is realized, and the power transmission line mountain fire prevention work can be effectively guided; according to a prediction conclusion, a corresponding emergency scheme can be made in time, the mountain fire can be handled in advance, and loss caused by the mountain fire to places nearby a power transmission line is reduced.

The present invention relates to the field of meteorological disaster early warning of an electrical power system, and discloses an electric transmission line rainstorm disaster fine prediction methodand system based on an inverse algorithm which are used to refine a rainstorm disaster prediction result to a concrete electric transmission line. The method comprises the steps of: generating a rainfall lattice point prediction file comprising rainstorm disaster index prediction value of each lattice point through a rainstorm disaster prediction and calculation model, and employing the space search of a power grid GIS to perform space intersection for the electric transmission line tower coordinate information and obtain the electric transmission line tower information of a rainstorm area; employing a power grid GIS space line to form electric transmission line space information of a rainstorm disaster; and employing a meteorological factor space interpolation method and combining the electric transmission line of the rainstorm disaster influence to generate an electric power line rainstorm disaster prediction and early warning distribution map through the rainstorm disaster prediction lattice point file, employing the same color to describe the electric transmission lines in the same rainstorm disaster index, and employing other colors to distinguish the electric transmission lines in different rainfall indexes.

The invention discloses a glacial lake burst early-warning method. Aiming at the defects that the adopted index is not reasonable, modification on observation data is over-simple in the existing glacial lake burst early-warning method, the invention provides the glacial lake burst early-warning method comprehensively considering the affects from an atmospheric temperature and the rainfall condition. According to the method, long-term observation data of temperature and rainfall in an observatory station is obtained and analyzed by a control center, a cumulative positive temperature daily increasing speed value TV in early days of a forecast date and a rainfall daily increasing speed value RV in early 30 days of the forecast date are calculated, and the TV value and the RV value are simultaneously subjected to double-index comparison with a burst warning curve RV=2.7214TV<-0.956> to judge the burst risks of the glacial lake, and an alarming system is started by the control center if necessary. According to the method, the corresponding relationship between two specific indexes of the cumulated temperature daily increasing speed value and the rainfall daily increasing speed value in early 30 days which are selected from numerous temperature and rainfall indexes and the glacial lake burst danger is determined, and the burst early-warning curve is determined. The forecast is carried out according to a more comprehensive and reasonable climate index, so the early-warning effectiveness is increased.

The invention discloses a hydrological drought forecasting and early warning method based on meteorological drought transmission. A standard rainfall index is selected to represent the meteorologicaldrought, and the standardized runoff index is selected to represent the hydrological drought. The drought occurrence time, the drought duration time, the drought duration time, the average drought strength, and the drought intensity are selected to be used for analyzing the drought characteristic from the meteorological drought to the hydrological drought. Based on the calculated SSI index sequence and the SPI index sequence having 1-24 month time scale, the time scale corresponding to the SPI index having the most correlation with the SSI index is determined by adopting a Pearson correlationmethod, and is marked as SPIn. Based on the SSI index sequence and the SPIn index sequence, a response model from the hydrological drought to the meteorological drought is established. Based on a response model having the best fitting degree, the drought characteristic threshold range required by hydrological drought with different drought levels is analyzed and calculated. According to the intensity and duration information of meteorological drought, a plurality of large-strength and long-lasting hydrological drought can be judged in the future.

The invention discloses a design flood derivation method and system based on equal reliability in a climatic change scene. The invention belongs to the field of reservoir flood control safety design.A future climate change scene is obtained based on global climate mode output and a two-stage deviation correction method considering meteorological variable correlation; a VIC distributed hydrological model is driven to obtain a runoff situation considering the influence of climate change; a principal component analysis method is adopted to preferably select an extreme rainfall index influencingflood characteristics as a covariable; a time-varying hydrological frequency analysis model is constructed by considering inconsistency of a flood sequence, a flood design value combination under a climatic change scene is deduced based on an equal reliability method, a most possible value is adopted as a final design value, and a 95% confidence interval is selected to estimate uncertainty of thefinal design value. According to the method, the influence of climate change on flood characteristics can be fully considered, and an important reference basis with high operability can be provided for deducing adaptive design flood in a climate change scene.

The invention discloses a wheat scab monitoring method based on satellite remote sensing, and belongs to the technical field of remote sensing monitoring, and the method comprises the steps: firstly obtaining the remote sensing data and meteorological rainfall data of a wheat planting region in a predicted year and N years before the predicted year, wherein the remote sensing data comprise multispectral data and thermal infrared remote sensing data; calculating a relative growth vigor index by utilizing the multispectral data; obtaining a normalized surface temperature index by using the multispectral data and the thermal infrared remote sensing data; obtaining a normalized precipitation index by using the meteorological precipitation data; and finally, calculating the wheat scab incidenceindex of the predicted year by utilizing the relative growth vigor index, the normalized earth surface temperature index and the normalized rainfall index. According to the method, the interference of the difference of wheat growth vigor in different areas in large-scale monitoring on the gibberellic disease monitoring model can be well reduced, and the objectivity and practicability of gibberellic disease remote sensing monitoring are improved.

The invention discloses a mountain torrent disaster critical rainfall index calculation method and system. The mountain torrent disaster critical rainfall index calculation method comprises the steps:determining a characteristic time period of a mountain torrent disaster; carrying out sliding statistics on the excitation rainfall of each characteristic time period; calculating early-stage influence rainfall corresponding to the excitation rainfall in each characteristic time period; based on the excitation rainfall of each characteristic time period where the mountain torrent disaster occursand the early-stage influence rainfall corresponding to the excitation rainfall, constructing an excitation rainfall and early-stage effective rainfall state space of each characteristic time period of the mountain torrent disaster; based on the Fisher algorithm and the state space, establishing a critical relationship model of exciting rainfall and early-stage influencing rainfall in each characteristic time period of the mountain torrent disaster, obtaining a classification linear discriminant equation, and calculating a critical rainfall value of each characteristic time period according tothe classification linear discriminant equation; and inputting the excitation rainfall of each characteristic time period of the to-be-identified disaster prevention object and the early-stage influence rainfall corresponding to the excitation rainfall of each characteristic time period into the critical relationship model, and outputting an alarm or non-alarm result.

The invention discloses a critical rainfall estimation method based on rainfall spatial distribution. The critical rainfall estimation method comprises the steps that firstly, estimating the confluence flow velocity in each grid unit in a hydrological response unit secondly, on the basis of the confluence flow velocity in each grid unit, obtaining a confluence time grid in the drainage basin; performing spatial interpolation on the time-interval rainfall data observed by the station to obtain a rainfall spatial distribution grid, and combining the convergence time grid to obtain a time-interval dynamic rainfall index; counting an accumulated dynamic rainfall index before a flood peak occurs in the flood process; then counting the tension water content saturation, namely the initial soil water content saturation, of the day before the flood process begins; finally, whether the flood is an over-alarm flood is judged, and making a cumulative dynamic rainfall indicator - initial soil watercontent saturation scattering point graph on the basis of multiple floods and i, marking points corresponding to the overwanted flood, and drawing a critical rainfall line in combination with the points corresponding to the overwanted flood.

The invention discloses a rainfall-type geological hazard warning and prediction model based on RBFN machine learning and a learning method thereof, belonging to a geological hazard monitoring technology. The model comprises an input layer, a hidden layer and an output layer. The input layer is a vector X=(x1,x2), wherein x1 is soil raining amount indix and x2 is hour raining amount. The output layer is a scalar RBFN value; The hidden layer is composed of m radial basis functions. The unit of operation of an array composed of hidden layers is the node of the hidden layers, which contains a central vector c and has the same dimension as the parameter vector x of the input layer. With taking rainfall as the main inducement of disasters, rainfall and actual disasters as a priori events, a rainfall-type geological hazard early warning and prediction model is established. The RBFN model is used to classify the early warning units by introducing soil rainfall index and machine learning algorithm into dimensionless soft thresholds, which avoids the drawbacks of subjective and statistical methods in the past.

The invention provides a data processing method and device applied to rainfall forecast, equipment and a medium. The data processing method applied to rainfall forecast comprises the following steps: determining an inversion rainfall rule according to historical rainfall data, and determining at least one rainfall index according to current rainfall data; based on an inversion rainfall rule, obtaining at least one piece of current inversion rainfall data corresponding to the at least one rainfall index one by one; and processing the current inversion rainfall data, and determining average rainfall forecast data corresponding to the current rainfall data. Therefore, by means of the terrain following coordinate system objective rainfall forecasting method based on at least one rainfall index numerical forecasting, the numerical model forecasting potential can be brought into full play, the accuracy of numerical model rainfall forecasting can be improved, and the rainfall forecasting level or capacity of mountainous areas and drainage basins with steep terrains can be improved.

The invention discloses an animal husbandry drought loss dynamic assessment method, which comprises the following steps: collecting data and preprocessing the data to drive an ecosystem model and calibrate the ecosystem model; calculating a standard rainfall index, and performing drought identification by using the standard rainfall index; calculating a difference value between the average herbageyield of the normal year and the average herbage yield of the drought year, and evaluating the influence of drought events of different grades on the herbage yield; evaluating the influence of drought on the animal carrying capacity through quantitative conversion and equivalent replacement of the pasture-animal carrying capacity, calculating the number of lost sheep units, and evaluating the economic loss value of a drought disaster according to the price of the sheep units; and constructing a drought loss assessment dynamic model and carrying out dynamic assessment on the drought loss. Byadopting the technical scheme of the invention, the influence of the drought on the grassland productivity can be accurately evaluated, the method has higher precision, sensitivity and applicability,and the animal husbandry economic loss caused by the drought can be quickly evaluated.

The invention discloses a method for analyzing correlativity between an ENSO index and rainfall based on factorial design, and belongs to the technical field of climate remote correlation. The method comprises the following steps: 1, collecting an ENSO index and rainfall data of a research area; 2, processing rainfall data of the research area into a monthly mean value, a monthly extreme value and a monthly abnormal value rainfall matrix along a time sequence and a spatial position; 3, calculating the influence of each factor and the interaction between the factors on the response variable, screening out the factor with the maximum influence and the factor level, and determining a factor combination with the most significant correlation between the ENSO index and the rainfall index of the specific region; 4, arranging the ENSO indexes into a matrix which is in one-to-one correspondence with the precipitation index time sequence and is 1-1 month earlier than the precipitation indexes; and 5, checking the correlation through a t-test formula. According to the method, the problem of uncertainty between the ENSO index and the rainfall correlativity of different regions is solved; and the influence of interaction between different factors on the correlativity can be analyzed.

The invention discloses a satellite-borne detection data global ocean rainfall type discrimination device and method. The device comprises an ocean rainfall pixel discrimination module based on microwave imager ocean rainfall product data, a data quality control module in the ocean area based on satellite-borne microwave data, a rainfall type rain index value calculation module for observing brightness temperature based on a microwave imager, a convection rain probability calculation module in a microwave imager pixel, a data processing module for a rainfall index value data set and a tropospheric rainfall probability data set, and a rainfall type discrimination module in a microwave imager observation pixel. The influence factors of the microwave imager for observing the ocean surface rainfall type discrimination are disclosed, a new method is provided for research and development of business products such as rainfall types of the FY-3 microwave imager in China, and a foundation is laid for improving the subsequent rainfall inversion precision.

The invention relates to the technical field of power transmission and distribution, and discloses a logistic regression-based power grid maintenance index calculation method and system, which are used for realizing calculation of a power grid maintenance index based on meteorological elements and providing a guarantee for reasonable arrangement of power grid maintenance. The method comprises thesteps of obtaining basic information of a to-be-overhauled power transmission line, establishing a basic database, and establishing a weather database of precipitation day by day in an area where theto-be-overhauled power transmission line is located; predicting the daily maximum temperature and relative humidity of the area where the power transmission line to be overhauled is located and the daily precipitation of seven days in the future by adopting a numerical mode; carrying out grid division on the region where the power transmission line to be overhauled is located according to latitudeand longitude, and corresponding to the numerical mode result; defining a high-temperature index, a rainfall index and an icing index, calculating a power grid maintenance probability according to the high-temperature index, the rainfall index and the icing index, determining the power grid maintenance index according to the power grid maintenance probability, and judging whether maintenance is needed according to the power grid maintenance index.

The invention provides a rainfall sediment transport model construction method suitable for multiple landform type areas, and the method comprises the steps: analyzing rainfall characteristic indexes and subarea sediment transport of different landform type areas of a watershed according to rainfall data of a rainfall station of the watershed, actually measured sediment transport data of a hydrometric station and area data of the different landform type areas; adopting a non-parametric statistical test method to analyze the modeling representative time period, obtaining an analysis result, constructing single-index rainfall sediment transport models of different landform type areas according to the analysis result, the rainfall characteristic indexes of the different landform type areas of the watershed and the subarea sediment transport amount, analyzing, judging and screening the effective rainfall indexes, constructing a partition multi-index rainfall sediment transport model based on the screened effective rainfall indexes, carrying out quantitative constraint on a value range of parameters in the partition multi-index rainfall sediment transport model to obtain a partition optimal model, calculating the sediment transport amount of the whole watershed by adopting the partition optimal model, and analyzing the rationality of the partition optimal model according to the sediment transport amount of the whole watershed. Operation is easy, and results are reliable.

The invention discloses a method for determining a regional atmospheric precipitation pattern index R, belongs to the technical field of atmospheric precipitation index analysis, and can solve the problem that the existing precipitation pattern index cannot reflect the change of spatial and temporal distribution of precipitation and cannot reflect the regional air water vapor resource quantity and supply condition. The determination method comprises the following steps: obtaining rainfall data of a to-be-measured time period of a research area, and obtaining an area rainfall index M according to the rainfall data; obtaining the total water vapor transport flux Q of the whole atmosphere of a unit air column of the research area, and obtaining an area water vapor transport index T according to the total water vapor transport flux Q; and according to the regional rainfall index M and the regional water vapor transport index T, calculating to obtain a regional atmospheric rainfall pattern index R.

The invention discloses a mountain torrent disaster zoning method based on machine learning. The mountain torrent disaster zoning method comprises the following steps: (1) extracting various short-time heavy rainfall indexes and underlying surface data indexes; (2) screening key factors of mountain torrent disasters from the short-time heavy rainfall index and the underlying surface data index; (3) constructing a self-organizing mapping neural network model (SOM) to perform primary spatial clustering on the mountain torrent disasters, and then performing secondary clustering; (4) comprehensively evaluating the secondary clustering result of the mountain torrent disasters by using the external indexes and the internal indexes, and determining an optimal clustering scheme of the mountain torrent disasters; and (5) calculating the standard deviation rate of the area of the zoning pattern spots, merging and processing the tiny patches, and obtaining a final mountain torrent disaster zoning result. According to the method, key factors related to mountain torrent disasters are selected through a random forest model, then two-stage hybrid clustering is carried out by utilizing SOM, and mountain torrent disaster space division is realized by combining various machine learning methods.

The invention discloses a rainfall inversion method and system based on a sunflower stationary satellite. The method comprises the following steps: pre-selecting a large amount of satellite channel data and elevation data as a first training sample set, and selecting a large amount of site live rainfall statistics as a second training sample set; inputting the first sample set and the second sample set into a convolutional neural network for training, and obtaining a weather model set and a rainfall intensity model set through a three-layer convolution classifier and a residual neural network regression device; and satellite channel data and elevation data of a current station are collected, the current satellite channel data and elevation data are input into the sun and rain model set, whether the current station is a rainfall area or a non-rainfall area is determined, and for the rainfall area, the rainfall statistic of the rainfall area is input into the rainfall intensity model, namely, a satellite inversion live rainfall index is output. According to the method of combining the rain and sun classification and the rainfall intensity model based on the neural network, the accuracy of rainfall inversion is improved.

The invention provides a method and device for determining the drought grade of a grassland, electronic equipment and a medium. The method comprises the following steps: acquiring meteorological state data, soil state data and vegetation state data of a region corresponding to each meteorological station in a plurality of meteorological stations in a target grassland; according to various state data, determining a relative water consumption index, a standardized rainfall index, a soil relative humidity index and a temperature vegetation drought index of a corresponding area of each meteorological station; according to each index, the seasonal adjustment coefficient and the weight corresponding to each index, determining a comprehensive drought index of the region corresponding to each weather station; determining a comprehensive drought index of other areas in the target grassland by using a spatial interpolation algorithm; and determining the drought level of each region in the target grassland based on the comprehensive drought index of each region and a preset corresponding relationship. The method can comprehensively consider meteorological, soil and vegetation factors to determine the drought grade of the grassland, can more accurately and comprehensively judge the drought condition, and is more suitable for the grassland environment.