Cloudy with a Chance of Landslides

Assessing Landslide Characteristics and Developing Susceptibility and Exposure Maps in Malawi


Team: Maggi Klug, Jordan Cissell, Max Grossman

Who am I?

Geography & International Affairs Student at GW

Who they are.

Jordan is a masters Geography student at University of Alabama, and Maggi is an undergraduate Atmospheric Sciences student at UAH

Where did I study landslides?

Hunstville, Alabama at the National Space Science Technology Center for NASA DEVELOP

Project Partners

NASA SERVIR Applied Sciences Team & Coordination Office

The Regional Center for Mapping of Resources for Development

Malawi

Topography, climate, seismotology, agriculture, and energy.

Disaster mitigation and management context

Disasters~development

Study Objectives

Update Global Landslide Catalog

Generate Susceptibility and Exposure Maps

Conduct Rainfall Assessment

Earth Observations

TRMM-PR

GPM-IMERG

Landsat 5-TM & Landsat 8-OLI

SRTM-v3 C-band

Global Landslide Catalog

Google Earth

Historical Slider

Time

Source: Google Earth

Landslide Susceptibility & Exposure Maps

Data Aquisition

Create topographic, land cover, soil type, and other landslide related datasets and landslide point layer.

Data Processing

Create Susceptibility Map using the Maximum Entropy Model, and intersect Susceptibility Map with LandScan to create Exposure Map.

Accuracy Assessment

Sample 25% of landslide points, consider “true” if susceptibility above 0.50. “True” points divided by all sampled points is accuracy.

Map Results


Susceptibility in northwest, central, and southeastern Malawi.

Source: NASA DEVELOP


Exposed cities include Blantyre, Zomba, and Mzuzu.

13 million people are in susceptible areas.

Rainfall Assessment

Data Aquisition

Download TRMM, GPM, and CHIRPS 2.0 at monthly accumulation and average daily precipitation per month.

Data Processing

Extract values at 500 random points for all datasets. Subset upper quartile for each dataset.

Data Analysis

Calculate and report correlation coefficient for GPM & CHIRPS 2.0 and TRMM & CHIRPS 2.0.

Rainfall Assessment Results

Data Source Temporal Resolution Avg. Correlation Min. Correlation Max Correlation
TRMM
Tot. Month
Avg. Daily
0.358
0.373
-0.198
-0.058
0.890
0.866
GPM
Tot. Month
Avg Daily
0.199
-0.025
-0.117
-0.538
0.745
0.607
Study Period: March 2014 - January 2015

What I learned

Modeling Landslides in Challenging

Preparing Data is very important

Teamwork is crucial

Future work

Test model at finer scales

Incorporate Inundation Duration Curves

Thanks!

Contact: m.grossman1813@gmail.com

Site: maxgrossman.github.io

All opinions, findings, and conclusions or recommendations expressed are mine and do not necessarily reflect the views of the National Aeronautics and Space Administration.