Every year, landslides–that the motion of rock, soil, and debris down a slope–cause tens of thousands of deaths, billions of dollars in damages, and disruptions to roads and power lines. Since terrain, characteristics of the rocks and soil, weather, and climate all contribute to landslide activity, accurately pinpointing areas most at risk of those dangers at any given time can be a challenge. Early warning systems are usually regional–based on region-specific data supplied by ground detectors, field observations, and rain totals. However, what if we can identify at-risk areas anywhere in the world at any time?
Enter NASA’s Global Landslide Hazard Assessment (LHASA) version and mapping instrument.
LHASA Version 2, published last month together with corresponding research, is a machine-learning-based model that assesses a collection of individual variables and satellite-derived datasets to make customizable”nowcasts.” These timely and concentrated nowcasts are estimates of possible landslide activity in near-real time for every single 1-square-kilometer place between the poles. The model factors in the incline of the land (higher slopes are more vulnerable to landslides), distance to geologic faults, the makeup of stone, past and present rainfall, and satellite-derived soil rain and snow bulk information.
“The model processes all of this data and outputs a probabilistic estimate of landslide hazard in the form of an interactive map,” said Thomas Stanley, Universities Space Research Association Laboratory in NASA’s Goddard Space Flight Center at Greenbelt, Maryland, who led the research. “This is valuable because it provides a relative scale of landslide hazard, rather than just saying there is or is not landslide risk. Users can define their area of interest and adjust the categories and probability threshold to suit their needs.”
In order to”teach” the model, researchers entered a table with each of the relevant landslide factors and many locations that have recorded landslides in the past. The machine learning algorithm takes the table and tests out different potential scenarios and outcomes, and when it finds the one which matches the data most accurately, it sparks a decision tree. It then identifies the errors in the decision tree and calculates another tree that fixes those errors. This process continues until the model has”learned” and improved 300 times.
“The result is that this version of the model is roughly twice as accurate as the first version of the model, making it the most accurate global nowcasting tool available,” said Stanley. “While the accuracy is highest–often 100%–for major landslide events triggered by tropical cyclones, it improved significantly across all inventories.”
Version 1, released in 2018, was not a machine learning model. It combined satellite precipitation data using a worldwide landslide susceptibility map to create its nowcasts. It made its predictions with one choice tree largely based on rainfall data from the preceding week and categorized every grid cell as low, medium, or high risk.
“In this new version, we have 300 trees of better and better information compared with the first version, which was based on just one decision tree,” Stanley said. “Version 2 also incorporates more variables than its predecessor, including soil moisture and snow mass data.”
Generally speaking, soil can only absorb so much water before getting saturated, and blended with different conditions, posing a landslide risk. By incorporating soil moisture data, the model can differentiate how much water is currently present in the soil and how much further rain would push it past that threshold. Likewise, if the mold knows the total amount of snow within a given area, it can factor in the additional water entering the soil as the snow melts. This data comes in the Soil Moisture Active Passive (SMAP) satellite, which will be managed by NASA’s Jet Propulsion Laboratory at Southern California. It launched in 2015 and provides continuous soil moisture coverage.
LHASA Version 2 also adds a brand new exposure feature that assesses the supply of roads and people in each grid cell to calculate the number of individuals or infrastructure exposed to landslide hazards. The exposure data is downloadable and was integrated into the map. Adding This Kind of information about vulnerable roads and inhabitants vulnerable to landslid