CN102005022A - Method for comprehensively evaluating casualty loss of link disaster by remote sensing - Google Patents

Abstract

The invention discloses a method for comprehensively evaluating the casualty loss of a link disaster by remote sensing. Aiming at the defects of too single disaster type, too diverse evaluation objectives and low timeliness of disaster damage evaluation in the conventional disaster monitoring and evaluation technology, the invention provides a technique capable of monitoring geologic disasters such as flooding and landslide by remote sensing, rapidly recognizing a damage target and evaluating the casualty loss of a plurality of disasters, which contributes to rapid evaluation of a plurality of disaster types, a large amount of data and a plurality of disaster-bearing bodies. Based on the technique for evaluating the casualty loss of the link catastrophe disaster by remote sensing, the method is characterized by comprising the following four units: a disaster target unit, a disaster information extracting unit, a casualty loss model unit and a casualty loss statistical unit. The technique for evaluating the casualty loss by remote sensing consists of the four units and can effectively realize rapid evaluation on the casualty loss of the plurality of disaster types caused by great natural catastrophes.

Description

The casualty loss remote sensing comprehensive estimation method of chain disaster

Technical field

Belong to disaster remote sensing monitoring and assessment technology field, specially the casualty loss evaluation of the disaster chain that causes at catastrophe realizes that the calamity of polytype disaster is decreased assessment technology.

Background technology

Along with the develop rapidly of China's economy and society, the density of population increases, and infrastructure is sharply expanded, and makes that human life and property loss that disaster causes are even more serious.And huge disaster, regular meeting brings out the generation in succession of secondary disaster, forms the disaster chain, and the breakdown strength of chain type disaster is much larger than the simple superposition of various single disaster types, and overall heavy losses can work the mischief.Directly caused great earthquake as 5.12 Wenchuan special violent earthquake catastrophe chains, and the landslide of causing, chip stream grade dried rhizome of rehmannia matter disaster, caused large-scale checked-up lake, form flood, flood residential area, road, farmland, cause casualties.Chain effect causes the stack of multiple disaster type, has caused the rescue measure to be difficult to carry out, time delays, the major disaster loss of the overall situation that works the mischief.For China, hydrometeorological huge disaster chain, i.e. " heavy rain-flood-landslide, rubble flow " has the frequency of generation height, and disaster intensity is big, the disaster chain is got in touch characteristics closely.Huge hydrometeorological disaster directly forms precipitation, causes basin-wide flood, and cause coming down, the generation of rubble flow grade dried rhizome of rehmannia matter disaster, form the catastrophe chain.At present, domestic calamity is decreased to estimate and is primarily aimed at single flood type, can't realize towards the calamity damage assessment of the multiple secondary disaster type of catastrophe initiation.Moreover, calamity at multiple disaster type is decreased assessment, some that have realized software, data, model and method abroad are tentatively integrated, but choosing of casualty loss target, need detailed field survey data support, be difficult to realize the quick calamity damage assessment of multiple disaster type, can not satisfy the timeliness demand of disaster assessment.

Deficiency at prior art and method, the comprehensive casualty loss assessment technique method of data such as we have developed and have a kind ofly fully utilized that the basis is geographical, resource environment and socioeconomic statistics data and disaster remote sensing monitoring data, can carry out calamity simultaneously and decrease Target Recognition, carry out the disaster-stricken quick casualty loss assessment technology method of town dweller's land used, farmland (arable land), road and population of multiple disaster type in the chain disaster simultaneously.

Summary of the invention

The present invention is directed in the existing disaster monitoring assessment technology, the disaster type is too single, evaluation objective is too various and calamity is decreased the deficiency of estimating poor in timeliness, a kind of remote sensing monitoring that can carry out flood, landslide geologic hazard simultaneously is provided, and the quick identification calamity assessment technique method of decreasing target, carrying out multiple hazard-affected body casualty loss, be convenient to realize the rapid evaluation of disaster-ridden kind, big data quantity, many hazard-affected bodies.

Technical solution of the present invention is as follows:

Based on the casualty loss remote sensing assessment technology of chain disaster, it is characterized in that comprising following four unit: disaster object element, the condition of a disaster information extraction unit, casualty loss model unit, casualty loss statistic unit, totally four unit.By the casualty loss remote sensing assessment technology method that above-mentioned four unit are formed jointly, can realize effectively that major natural disasters cause the quick calamity damage assessment of multiple disaster type.

First module, described disaster object element, i.e. the production of casualty loss evaluation objective data, mainly comprise the people or with the evaluation objective of relating to persons, specifically comprise town dweller's land used data, road data, farmland (arable land) data and spatialization demographic data, the production of four class hazard-affected body data.

Unit second, promptly described the condition of a disaster information extraction unit comprises: the identification and the extraction of water body scope extraction model and sliding mass rubble flow.In the condition of a disaster information extraction unit of flood,, extract water body scope and area according to remote optical sensing feature, the radar remote sensing feature of water body; In the condition of a disaster information model of landslide disaster,, extract landslide scope or rubble flow scope and scale parameter thereof according to the remote Sensing Interpretation feature of sliding mass, rubble flow.

Unit the 3rd, promptly described casualty loss model unit mainly comprises town dweller's land used casualty loss model, farmland (arable land) casualty loss model, road casualty loss model, the disaster-stricken model of population.Mainly comprise town dweller's land used flood loss model, town dweller's land used landslide casualty loss model, farmland (arable land) flood loss model, farmland (arable land) landslide casualty loss model, road flood loss model, road landslide casualty loss model, flood population suffered from disaster model, landslide population suffered from disaster model.

Unit the 4th, promptly described casualty loss computing unit mainly comprises the casualty loss situation of calculating different hazard-affected bodies; Stack administrative division data are carried out the subregion of casualty loss and are added up.

Compared with prior art, the present invention has following characteristics:

A) the multiple disaster type monitoring requirements that causes towards catastrophe is carried out the casualty loss evaluation of multiple disaster type simultaneously, can satisfy the information requirement of different industries department.

B) given prominence to the key points the people or with the ground object target of relating to persons, make the casualty loss evaluation objective be more suitable for the great casualty loss evaluation of the chain type disaster that catastrophe causes.

C) abandoned the shortcoming that existing calamity is decreased assessment technique input parameter time and effort consuming, realized multiple hazard-affected body (urban settlement, farmland, road, population suffered from disaster) multiple cause the calamity characteristic element (depth of water, flood last, the landslide influence) the simplification of sxemiquantitative casualty loss function and integrated, improved the disaster monitoring timeliness.

Description of drawings

The techniqueflow chart of the casualty loss remote sensing comprehensive estimation method of Fig. 1 chain disaster.

The casualty loss assessment models structural drawing of Fig. 2 town dweller's land used.

The casualty loss assessment models structural drawing in Fig. 3 farmland (arable land).

The casualty loss assessment models structural drawing of Fig. 4 road.

Fig. 5 population suffered from disaster's disaster assessment models structural drawing.

Embodiment

The present invention utilize geo-spatial data, statistics and remotely-sensed data carry out catastrophe chain disaster casualty loss remote sensing comprehensive estimation method techniqueflow chart as shown in Figure 1.Utilize this method, cause multiple disaster, can carry out the rapid integrated assessment of catastrophe chain casualty loss towards the catastrophe chain effect.Concrete enforcement is as follows:

1. disaster object element

Casualty loss evaluation objective data: adopt the man-machine interaction decomposition method,, discern and hazard-affected bodies such as the farmland of sketching (arable land), road, town dweller's land used according to the spectrum and the texture information of middle and high resolution data; Demographic data is mainly utilized statistics, land use pattern, light data, realizes the spatialization of demographic data.Casualty loss evaluation objective data are as the basic data of casualty loss assessment.

◆ urban settlement identification: texture shows as culture's features such as regular rectangle, square; Spectral reflectivity is extremely stable, seasonal variations can occur hardly; High-resolution data very is easy to identification;

◆ farmland (arable land) identification: in easy flights in 25 degree; Have comparatively regular border on the image texture or have the terraced fields border;

◆ road Identification: linear trace is remarkable, shows as tangible artificial trace; Multi-link between the residential area, villages and small towns; High-resolution data very is easy to identification;

◆ demographic data:

Based on the consensus data, calculate the demographic data of space distribution:

$p=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{ax}}_j;$

Wherein, p is the size of population of a certain area of space; I is the urban settlement number; A is town dweller's dot factor (between 0-100000); X accounts for the ratio of territory, whole county urban settlement area for this urban settlement.

2. the condition of a disaster information extraction unit

Mainly comprise the extraction of water body scope, landslide, and the investigation of bathymetric data, landslide thickness data.

◆ water body extracts

Mainly utilize man-machine interaction method, extract the water body scope, form water body scope raster data, water body DN=1, non-water body DN=0.The water body of high-definition remote sensing data extracts precision and reaches more than 90%.

◆ landslide extracts

The main method that adopts man-machine interaction, extraction landslide, rubble flow cause the calamity scope, and formation landslide, rubble flow cause calamity scope vector data.Behind the vector rasterizing, landslide, rubble flow cause calamity scope DN=1, non-territory, the disaster area DN=0 that causes.The landslide of high resolving power experimental data, rubble flow cause calamity scope extraction precision and reach more than 95%.

3. casualty loss model unit

◆ the casualty loss assessment models of town dweller's land used

The casualty loss assessment models of town dweller's land used can be divided into flood and landslide casualty loss model (Fig. 2).

The flood loss model of town dweller's land used:

The disaster area of town dweller's land used is an important evaluation index of assessment flood loss.In town dweller's land used, suffer flood inundation on tracks, think then and stand the flood loss that the model concrete form is as follows:

${D\_\mathrm{Urbanland}}_{\mathrm{Flood}}=\left\{\begin{array}{cc}0& \mathrm{IsFlood}=0\\ x_F& \mathrm{IsFlood}=1\end{array}\right.$

Wherein, D_Urbanland _FloodTown dweller's land area that expression is influenced by flood; x _FExpression is subjected to the residential area area of flood influence.

Town dweller's land used landslide casualty loss model:

Landslide debris fails to be convened for lack of a quorum house and auxiliary facility is caused expendable influence, and in landslide residential area damage assessment models, the parameter that needs comprises that landslide debris distributions and residential area distribute.The residential area that is subjected to the landslide influence is defined as damage, and the model concrete form is as follows.

${D\_\mathrm{Urbanland}}_{\mathrm{Landslide}}=\left\{\begin{array}{cc}0& \mathrm{IsLandslide}=0\\ x_L& \mathrm{IsLandslide}=1\end{array}\right.$

Wherein, D_Urbanland _LandslideTown dweller's land area that expression is influenced by the landslide disaster; X wherein _LThe residential area area that expression is influenced by landslide.

◆ farmland (arable land) casualty loss model

The casualty loss assessment models in farmland (arable land), the casualty loss that is primarily aimed at the crops in the farmland (arable land) is assessed, and can be divided into flood and landslide casualty loss model (Fig. 3).

Farmland (arable land) flood loss model:

Crops flood loss percentage, not only relevant with species, growth period, and with the depth of water, flood and last close relation.Flood crop loss rate model has mainly been considered the depth of water and has flooded to last influence, the computing formula of definition:

During in depth of water h≤0.5 meter:

${D\_\mathrm{Farmland}}_{\mathrm{Flood}}=\left\{\begin{array}{cc}0.25& 1\&le;t<2\\ 0.31& 2\&le;t<3\\ 0.38& 3\&le;t<4\\ 0.46& 4<t<5\\ 0.53& 5\&le;t<6\\ 0.58& 6\&le;t<7\\ 0.63& t\&GreaterEqual;7\end{array}\right.$

During in the depth of water 0.5＜h≤1.0 meter:

${D\_\mathrm{Farmland}}_{\mathrm{Flood}}=\left\{\begin{array}{cc}0.36& 1\&le;t<2\\ 0.42& 2\&le;t<3\\ 0.48& 3\&le;t<4\\ 0.57& 4<t<5\\ 0.67& 5\&le;t<6\\ 0.76& 6\&le;t<7\\ 0.85& t\&GreaterEqual;7\end{array}\right.$

During in depth of water h＞1.0 meter:

${D\_\mathrm{Farmland}}_{\mathrm{Flood}}=\left\{\begin{array}{cc}0.47& 1\&le;t<2\\ 0.55& 2\&le;t<3\\ 0.64& 3\&le;t<4\\ 0.72& 4<t<5\\ 0.81& 5\&le;t<6\\ 0.87& 6\&le;t<7\\ 0.92& t\&GreaterEqual;7\end{array}\right.$

Wherein, D_Farmland _FloodThe arable land crop loss that expression is influenced by flood; H represents the depth of water (rice); T represent to flood and last (my god).

Lacking the bathymetric data support, only having and flood when lasting data, crops flood loss percentage computing formula:

${D\_\mathrm{Farmland}}_{\mathrm{Flood}}=\left\{\begin{array}{cc}0.1& 1<t<3\\ 0.3& 3\&le;t<7\\ 0.8& t\&GreaterEqual;7\end{array}\right.$

Wherein, D_Farmland _FloodThe arable land crop loss rate that expression is influenced by flood; T representative flood and last (my god).

Can obtain the spatial distribution state of flood crop loss rate by above-mentioned model,,, be defined as disaster-stricken farmland because of calamity causes damage according to the relevant criterion of Ministry of Civil Affairs; Be lost in and be defined as the farmland of causing disaster between 3 one-tenth to 8 one-tenth; Be defined as the total crop failure farmland more than being lost in 8 one-tenth.So promptly obtained flood crop loss product.

The landslide casualty loss model in farmland (arable land):

The landslide cliff debris is extremely serious for the influence of crops.Generally speaking, cliff debris is buried and can be caused crops total crop failure (100% loss), buries serious zone even can cause arable land can't reclaim (damage of ploughing).The crop loss model that landslide caused can be defined as:

${D\_\mathrm{Farmland}}_{\mathrm{Landslide}}=\left\{\begin{array}{cc}0& \mathrm{IsLandslide}=0\\ 1.0& \mathrm{IsLandslide}=1\end{array}\right.$

D_Farmland _LandslideThe arable land crop loss rate (IsLandslide=0 represents to plough landslide does not take place, and IsLandslide=1 represents that landslide takes place in the arable land) that expression is influenced by landslide

◆ road casualty loss model

The casualty loss assessment models of road can be divided into flood and landslide casualty loss model (Fig. 4).

The road casualty loss model of landslide:

${D\_\mathrm{Road}}_L=\left\{\begin{array}{cc}0& \mathrm{IsLandslide}=0\\ R_L& \mathrm{IsLandslide}=1\end{array}\right.$

D_Road _L, (IsLandslide=0 represents that landslide does not take place in this highway section, IsLandslide=1 represents that landslide, R take place in this highway section to represent the link length that influenced by the landslide disaster _LExpression is subjected to landslide to influence road section length)

The road casualty loss model of flood:

${D\_\mathrm{Road}}_F=\left\{\begin{array}{cc}0& \mathrm{IsFlood}=0\\ R_F& \mathrm{IsFlood}=1\end{array}\right.$

D_Road _F, (IsFlood=0 represents that flood does not take place in this highway section, IsFlood=1 represents that flood inundation on tracks, R take place in this highway section to represent the link length that influenced by flood _FExpression is subjected to the flood influence road section length)

◆ the population model of suffering from disaster

The casualty loss assessment models of population can be divided into flood and landslide casualty loss model (Fig. 5).

Flood population suffered from disaster model:

${D\_\mathrm{Population}}_F=\left\{\begin{array}{cc}0& \mathrm{IsFlood}=0\\ P_F& \mathrm{IsFlood}=1\end{array}\right.$

D_Population _L, representative is subjected to the size of population that flood influences, and (IsFlood=0 represents that this regional population is not subjected to flood, IsFlood=1 to represent that this regional population stands that flood damage influences, P _FExpression is subjected to flood damage population)

Landslide population suffered from disaster model:

${D\_\mathrm{Population}}_L=\left\{\begin{array}{cc}0& \mathrm{IsLandslide}=0\\ P_L& \mathrm{IsLandslide}=1\end{array}\right.$

D_Population _L, representative is subjected to the size of population that the landslide disaster influences, and (IsLandslide=0 represents that this regional population is not subjected to that landslide influences, IsLandslide=1 represents that this regional population stands that landslide influences, P _LExpression is subjected to landslide disaster population).

4. casualty loss computing unit

At the disaster evaluation objective, the condition of a disaster data that stack is extracted, call calamity and decrease evaluation model, the disaster-stricken length of flood of the flood loss of the flood disaster area of calculating town dweller land used, town dweller's land used landslide disaster disaster area, farmland (arable land) crops, the landslide casualty loss of farmland (arable land) crops, road, the disaster-stricken length of landslide disaster of road, population suffered from disaster's quantity that flood causes, population suffered from disaster's quantity that the landslide disaster causes.

By stack administrative division data, can access the disaster-stricken statistics of evaluation objective that the different disaster types in the territory, different rows administrative division cause.

Claims (1)

1. One kind fully utilize that remotely-sensed data, basis are geographical, the chain catastrophe casualty loss remote sensing assessment technique method of multi-source datas such as resource environment and statistics, it is characterized in that this method comprises as lower unit: disaster object element, the condition of a disaster information extraction unit, casualty loss model unit, casualty loss computing unit, totally four unit.By the comprehensive casualty loss remote sensing of the multi-source data assessment technology method that above-mentioned four unit are formed jointly, can effectively realize the emergency monitoring and the assessment of calamity damage of major natural disasters.

   First module, promptly described disaster object element comprises town dweller's land used, road, farmland (arable land) and population, as suffering disaster, forms the identification of heavy losses hazard-affected body.

   Unit second, promptly described the condition of a disaster information extraction unit comprises obtaining of flood information extraction, landslide geologic hazard information.

   Unit the 3rd, promptly described casualty loss model unit comprises town dweller's land used casualty loss model, farmland (arable land) casualty loss model, road casualty loss model, the disaster-stricken model of population.

   Unit the 4th, promptly described casualty loss computing unit mainly utilizes hazard-affected body, the condition of a disaster data of said units, calls the casualty loss model, calculates also statistics casualty loss.

Images