Agricultural Research Management Information System

Home
Research Summary
Search
About Us
- ARMIS
- Brochure
Contact Us
Report
User Request
Data Input
Help
- Operation Manual
  - PDF
  - Video
- Program Area & Commodity
We have reached 37600 number of research entries at this moment.

- Logout

Research Detail

Home
Research
Detail

Event-based Landslide Susceptibility Mapping Using Weights of Evidence(woe) and Modified Frequency Ratio (mfr) Model: a Case Study of Rangamati District in Bangladesh

Shamima Ferdousi Sifa
Department of Disaster Science and Management, Faculty of Earth and Environmental Sciences, University of Dhaka, Dhaka, Bangladesh

Tonoy Mahmud
Department of Disaster Science and Management, Faculty of Earth and Environmental Sciences, University of Dhaka, Dhaka, Bangladesh

Maria Abdullah Tarin
Department of Disaster Science and Management, Faculty of Earth and Environmental Sciences, University of Dhaka, Dhaka, Bangladesh

Dewan Mohammad Enamul Haque
Department of Disaster Science and Management, Faculty of Earth and Environmental Sciences, University of Dhaka, Dhaka, Bangladesh

Abstract/ Executive Summary:

Landslide hazard of 2017 in Rangamati district had devastating impacts on development,thereby making landslide susceptibility mapping a prerequisite for disaster risk management.This study aims to map the future landslide susceptible areas by overlaying the lands lide inventory of 2017 with causative factor maps using WoE and MFR and compare their results to determine that statistical model describes the susceptibility of the landslide occurrence better than the other. The analysis shows that although both models define the spatial relationship of past landslides with the triggering factors in a same way but in case of mapping, MFR had overestimated the high and low susceptible areas and under estimated the moderate susceptible areas than WoE. When validated from success rate curve by plotting the percentage of landslide susceptibility index rank against the percentage of cumulative landslide occurrence, it shows that the WoE model describes the landslides better than the MFR model. About 20% of the high susceptible areas include 85% of the total landslide area in case of the WoE model but the MFR model includes only 20%. On the other hand, the WoE model describes that 30% highly susceptible area covers more than 99% of the total landslide area while MFR defines only 78%.

Keywords: Landslide; Susceptibility; Inventory; weights of evidence (WoE); Modified frequency ratio (MFR)

Location:

Start Date:

End Date:

Program Area: Risk Management in Agriculture

Non-commodity: Modeling

Objectives:

This study aims to map the future landslide susceptible areas using WoE and MFR and compare their results to determine which statistical model describes the susceptibility of the landslide occurrence better.

Methodology:

Landslide susceptibility mapping is carried by over-laying the landslide inventory with the factor maps using the WoE and MFR models, a statistical approach that shows how factors contribute to the landslide occurrence. Therefore, a landslide inventory consisting the location of past landslides is produced and combined with triggering factor maps by giving different weight values derived from each model. Past landslides were used to train the data because the“past and present are keys to the future”; therefore,the future landslide susceptible areas are estimated from the past landslides and the preconditioning factors that are responsible for the slope failure assuming that the causative factors will also initiate future landslides under similar conditions. The results of the models were later validated from the gradient of the susceptibility index rank and cumulative landslide occurrence curve. The inventory map was produced from the SAR image, which has a 10-m spatial resolution and the offset tracking was employed to two SAR single look complex (SLC) images sensed on 6 June 2017 and 18 June 2017, which gave a 12 days temporal window. To produce the factor maps, 12.5 m ALOS PALSAR digita elevation model (DEM) and 30 m Landsat 8 optical images were used. The rainfall data are collected from the Bangladesh Meteorological Department. Landslide inventory preparation The landslide inventory shows the previous occurrence of a landslide that has occurred in the study area, and this was prepared using SAR offset tracking technique.Offset tracking provides displacement information parallel to SAR satellite track (Azimuth offset) and the track perpendicular (range offset) (Fialko, Sandwell, Simons,& Rosen,2005). The window patch intensity is measured to find the motion or displacement between two images of the same area (Lu,2016). First, the master (pre event) and slave image (post-event) are co-registered using a digital elevation model (DEM). Afterward,to calculate the offset, a ground control point (GCP) grid is specified for the master image. The displacement is measured by matching frequency peaks of a patch of pixels in master and slave image. If two corresponding windows give the same frequency peaks, they are considered stable (Pathier et al.,2006). Each applied to offset a correlation value is computed, thus permitting calculation of the corresponding matching correlation surface (MCS), whose ith and jth elements for the generic pixel of range and azimuth coordinates (x, y)are the following. Validation of landslide susceptibility models The models are verified using success rate curve to see how well each model describes the susceptibility of the landslide occurrence (Chung & Fabbri,1999). To validate, 25% of the landslide distributions are separated at the start of the modeling and the remaining 75% are used to train the model. The test data are then combined with the final weighted map that is reclassified into 32 classes to compute percentage The values are then plotted in a graph of percentage of cumulative landslide occurrence against the percentage of susceptibility index rank to obtain a success rate curve (Jebur, Pradhan, Shafri, Yusoff, & Tehrany, 2015).

Source of Information: GEOLOGY, ECOLOGY, AND LANDSCAPES 2020, VOL. 4, NO. 3, 222–235

Article Link (Online Journal): https://doi.org/10.1080/24749508.2019.1619222

Funding Source:

Budget:

Total Budget (Tk.) :

Achievement/Findings:

Rangamati district in Bangladesh got severely affected by the landslide event of 2017. This large number of events had made the slope unstable, thereby increasing the possibility of slope failure in the area greater than before. The aim of the study was to produce the landslide susceptibility map of the area based on landslide inventory of 2017, using weight of evidence (WoE) and modified frequency ratio (MFR) models to find out which model explains the future susceptibility better. An event-based (2017) landslide inventory was produced by processing radar image through SAR offset tracking. The creation of inventory involves a knowledge-driven approach, which was later used in combination with the causal factors to map the future susceptible areas to see the spatial relationship of each factor with the landslide occurrence that is the factors that were used in the study was combined with previous landslides to determine the condition of each factor in past landslide occurrence. Later, these values were used to train the data set and derive the weight values to see the association of the factors and possibility of landslide occurrence. The factors were then combined to see how these factors that were responsible for past land-slide occurrence also increase the susceptibility for future landslide occurrence, assuming that“past and present are keys to the future.”The study area was then classified into high, moderate, low, and very low susceptibility classes, and comparison was made between the WoE and MFR models, describing how well each model define the classes. The results of the Landslide susceptibility map of each method were validated from the success rate curves. The percentage of landslide susceptibility index rank is plotted against the percentage of cumulative landslide occurrence to produce the success rate curve. About 20% of the high susceptible areas include 85% of the total landslide area in case of the WoE model but the MFR model includes only 20%. The WoE model describes that 30% highly susceptible area covers more than 99% of the total landslide area in the case of success rate, while MFR describes only 78% that is again lower than the WoE model. As WoE explains the results better than MFR, the model was later used to find out which upazila of the study area is at high landslide susceptible areas compared to the other. Rangamati Sadar upazila has been identified to have high susceptibility toward landslide hazard as 46% and 33% of its area fall under high and moderate susceptible classes, respectively. Kaukhali, Rajasthali, and Kaptai upazila have 23%, 25%, 8% high, and 28%, 33%, 10% moderate susceptible areas, respectively. The results obtained from the research, thereby, might help the policy makers to take appropriate mitigation measures to prevent the severity of the land-slide hazard consequences in the study area.

Knowledge Management: Journal