Geological disasters refer to natural geological and human natural disaster emergencies caused by the earth's internal dynamics, external dynamics or artificial geological dynamics during the natural geological evolution of the global crust. Due to natural effects such as precipitation and earthquakes, geological disasters occur frequently all over the world. In addition to landslide disasters in my country, there are also various geological disasters such as landslides, debris flows, and land subsidence, which are characterized by various types, wide distribution, and great harm. The risk assessment of geological hazards focuses on assessing the degree of geological hazards occurring in a certain stage in a certain region according to various influencing factors and regional selection. In order to predict and analyze the possibility of geological disasters occurring in a certain terrain unit in the future. According to the breeding and development mechanism of geological hazards, existing data and technologies, as well as practical application needs, evaluation objectives and research funds, etc., using appropriate methods, the risk of geological hazards in the research area can be evaluated and analyzed through models. So how to deeply understand the disaster risk assessment model? How to efficiently process hazard data? How to establish a feasible geological disaster risk assessment and post-disaster reconstruction plan for specific regions? This course will provide a set of methods and cases based on ArcGIS.
ArcGIS software has the functions of spatial data and attribute data input, editing, query, simple spatial analysis statistics, output, report, etc., which provides the possibility for the organic integration of multi-source data, and also facilitates the establishment of flexible analysis modules. The function of spatial analysis is one of the important reasons why GIS is widely used. Using GIS analysis technology, statistical analysis of various factors, information superposition and compounding, research on types of geological disasters, distribution regularity level and disaster loss degree, etc., using methods such as risk index to evaluate and map the status quo of geological disaster risk, will enable The risk assessment of geological disasters is more efficient and scientific, which provides strong support for the construction of geological disaster databases.
Combining project practice cases and scientific research papers to explain. Introductory chapter, the quick introduction of ArcGIS software and the acquisition and understanding of GIS data sources; method chapter, the method of extracting disaster-causing factors, the establishment method of disaster risk factor analysis index system and the construction method of disaster risk evaluation model; The application method in disaster reconstruction, the establishment and optimization method of disaster susceptibility evaluation model using machine learning in Python environment.
Further understand the formation mechanism and disaster mode of geological disasters; master the methods of GIS in disaster risk assessment from the aspects of spatial data processing, disaster-causing factor extraction, spatial analysis, risk assessment and cartographic analysis; in specific practical cases, learn Use the principles and technical methods of geological disaster risk assessment, and at the same time learn the application methods of GIS in post-disaster reconstruction planning and other fields to improve the application ability of GIS technology.
Original Link: Application of GIS in Geological Hazard Risk Assessment and Post-disaster Reconstruction
Chapter 1, Basic Concepts and Platform Introduction
1. Basic concepts
Types of geological hazards
The Developmental Characteristics and Distribution Rules of Geological Hazards
Hazard Characteristics of Geological Hazards
Analysis of Geological Conditions Pregnant in Geological Hazards
Inducing Factors and Forming Mechanism of Geological Disasters
2. GIS principle and ArcGIS platform introduction
l Introduction to GIS
l ArcGIS foundation
l Spatial data collection and organization
l Spatial reference
l Spatial data conversion and processing
l Data editing in ArcGIS
l Visual representation of geographic data
l Spatial analysis:
Digital Terrain Analysis
Overlay Analysis
distance mapping
Density Mapping
Statistical Analysis
rearrange
3D analysis
3. Python compilation environment configuration
lPython comes with editor IDLE to use
lAnaconda integrated environment installation and use
lPyCharm environment installation and use
Chapter Two, Data Acquisition and Preprocessing
1. Data type introduction
2. Point data acquisition and processing
l Acquisition and processing of statistical data of disaster points
l Acquisition and processing of meteorological station data
Point data processing of meteorological stations
Weather Data Acquisition
data collation
exploratory analysis
Data interpolation analysis
3. Acquisition and processing of vector data
l Acquisition of vector data such as roads, faults, and water systems
l Euclidean distance
l Kernel density analysis
l River network density analysis
4. Raster data acquisition and processing
l DEM, remote sensing images and other raster data acquisition
l Image stitching, cropping, masking and other processing
l NoData value processing
l How to unify the row and column numbers
5. NC data acquisition and processing
l Introduction to NC data
l NC data acquisition
l Model Builder
l How to transfer NC data to TIF?
6. Data acquisition and processing of remote sensing cloud computing platform
l Introduction to Remote Sensing Cloud Platform Data
l How to obtain data from the cloud platform?
l Data upload and download
l Introduction to Basic Functions
l Vegetation Index Extraction
l Land use data acquisition
7. Python data cleaning
l Introduction and installation of Python library
l read data
l Uniform number of rows and columns
l Missing value processing
l Correlation analysis/collinearity analysis
l Principal component analysis (PCA) dimensionality reduction
l Data standardization
l Generate feature set
Chapter 3. Model and Method of Geological Hazard Risk Assessment
1. Evaluation model and method of geological disaster susceptibility
Evaluation unit determined
Susceptibility evaluation index system
Susceptibility Evaluation Model
Determination of weight
2. Evaluation of landslide susceptibility
l Evaluation index system
Terrain: elevation, slope, gully density, relief, etc.
Landform: landform unit, micro-geomorphology, overall topography, etc.
Stratum lithology: lithological characteristics, rock thickness, rock genetic type, etc.
Geological structure: faults, folds, joint fissures, etc.
Earthquake: Intensity, dynamic peak acceleration, historical seismic activity, etc.
Engineering geology: regional crustal stability, bedrock burial depth, main bearing layer lithology, bearing capacity, rock and soil engineering geological division, etc.
l Extraction of common indicators
Factor extraction such as slope, slope type, elevation, terrain relief, fault zone distance, engineering geological rock group, slope structure, vegetation coverage, distance from water system, etc.
l Index factor correlation analysis
(1) Calculation and analysis of correlation coefficient
l Information volume of evaluation indicators
l Determination of the weight of evaluation indicators
l Analysis and mapping of landslide susceptibility assessment results
Landslide Susceptibility Composite Index
Susceptibility Classification
Graphical analysis of susceptibility evaluation results
2. Evaluation of collapse susceptibility
3. Evaluation of debris flow susceptibility
l Debris flow evaluation unit extraction
l Debris flow evaluation index
Severity of landslides, length ratio of sediment recharge along the course, debris flow accumulation activity at the gully mouth, longitudinal slope of valleys, degree of regional structural influence, vegetation coverage of watersheds, engineering geological rock formations, reserves of loose deposits along the gully, watershed area, watershed Relative height difference, degree of channel blockage, etc.
l Selection of typical debris flow evaluation indicators
l Evaluation factor weight determination
l Analysis and drawing of evaluation results of debris flow susceptibility
Calculation of Debris Flow Susceptibility Composite Index
Determination of Susceptibility Classification of Debris Flows
Evaluation results of debris flow susceptibility
4. Comprehensive evaluation of susceptibility to geological disasters
Comprehensive geological disaster susceptibility value = MAX [debris flow disaster susceptibility value, collapse disaster susceptibility value, landslide disaster susceptibility value]
Chapter IV. Risk Assessment of Geological Hazards
1. Risk assessment of geological disasters
2. Risk assessment of geological disasters
l Selection of risk assessment factors
Under the action of certain triggering factors, the possibility of geological disasters of a certain scale and type occurring in a certain period of time in a certain area.
The complexity of regional structures, the development of active faults, and seismic activity may all induce geological disasters; the induction of heavy rainfall, the frequency and scale of disasters will also increase the probability of geological disasters.
l Quantification of risk assessment factors
Quantification of risk factors for landslides
Count the number and area of disasters at all levels, and use the information volume calculation method to obtain the information volume values at all levels.
Debris flow risk assessment factor weight
l Risk assessment and result analysis
3. Assessment of vulnerability to geological hazards
l Analysis of vulnerability factors of geological hazards
population vulnerability
house building vulnerability
agricultural vulnerability
forestry vulnerability
Livestock vulnerability
road traffic vulnerability
water vulnerability
l Extraction of population vulnerability evaluation factors
Population Density Data Processing
Use population density data to quantify population vulnerability, based on the population numbers obtained from the statistical yearbooks of each administrative unit, combined with housing construction area data, to quantify the spatial distribution of population, and based on GIS grid analysis, the population per unit area is obtained. density.
vulnerability assignment
Population Vulnerability Factor Extraction
l Building vulnerability assessment
Data Processing of Building Area Density
Quantifying building vulnerability with building-building area density data, using building-building area data,
Based on the grid analysis of GIS, the area of the building area per unit area is obtained, that is, the density of the building area.
vulnerability assignment
Extraction of Building Vulnerability Factors
l Assessment of the vulnerability of traffic facilities
Acquisition of road data
Use the ArcGIS buffer analysis function to form the surface file of the road
Different types of roads for assignment
Analysis of Road Vulnerability Distribution Results
l Comprehensive vulnerability assessment
Combined Vulnerability Overlay Weight
Extraction and Analysis of Comprehensive Vulnerability Evaluation Results
4. Extraction and analysis of geological disaster risk assessment results
Chapter Five, GIS Application Practice in Post-disaster Reconstruction
1. Analysis of emergency rescue path planning
l Spatial analysis such as surface analysis, cost weight distance, and raster data distance mapping;
l Using the basic method of thematic map drawing, the emergency rescue route map for geological disasters in Maoxian County, Sichuan Province was made,
l Extraction and analysis of the best path
2. Site selection analysis for disaster recovery and reconstruction
l Determining the influencing factors of site selection
l Determine the weight of each impact factor
l Collect and process data for each influencing factor: terrain analysis, distance cartographic analysis, reclassification
l Site selection analysis for restoration and reconstruction
3. Analysis of ecological environment changes after the earthquake
Use the powerful data acquisition, data processing, data storage and management, spatial query and spatial analysis, visualization and other functions of this type of software to evaluate ecological environment changes.
Chapter 6, Thesis Writing and Reproduction
1. Analysis of the main points of thesis writing
2. Analysis of paper submission skills
3. Case analysis of the paper
Case: Research on Disaster Susceptibility Evaluation Using Machine Learning
4. Some results are reproduced:
think:
ü Is it necessary to normalize the features before training
ü Why should we deal with missing values (Nan values)
ü What will be the impact of the high correlation between the input features
ü What are the training set, test set and validation set; why should they be divided like this
ü What are hyperparameters
ü What is overfitting and how to avoid it
Model introduction:
ü Logistic regression model
ü Random forest model
ü Support vector machine model
Implementation plan:
1. Linear Probability Model - Logistic Regression
ü Introduction
ü Selection of connection function: Sigmoid function
ü Hazardous factor data set: data introduction; correlation analysis; logistic regression model prediction; sample precision analysis; classification confusion matrix
ü Precautions
Two, SVM support vector machine
ü Linear classifier
ü SVM-kernel method: Introduction to kernel method; sklearn's SVM kernel method
ü Parameter optimization and adjustment
ü SVM data set: support vector machine model prediction; sample accuracy analysis; classification confusion matrix
3. Python implementation of Random Forest
ü Dataset
ü Random selection of data
ü Random selection of features to be selected
ü Explanation of related concepts
ü Parameter optimization and adjustment: random forest decision tree depth tuning; CV cross-validation definition; confusion matrix; sample precision analysis
ü Implement Random Forest based on pandas and scikit-learn: data introduction; random forest model prediction; sample precision analysis; classification confusion matrix
4. Comparative Analysis of Methods
ü Precision Analysis
Comparative analysis of results
