Global Ground Movement Susceptibility Raster (GMSR) (Class 1–7)

Interactive map with scientific data analysis, point lookup, and detailed environmental information

Map Information

The Global Ground Movement Susceptibility Raster (GMSR) is a global geospatial dataset representing relative susceptibility to long-term ground movement hazards.

Data Source:
Environmental Data
Units:
Index (1-7)
Coverage:
CONTINENTAL
Citation:
Mazzella, J., Mazzella, N. (2026). Global Ground Movement Susceptibility Raster (GMSR) Class 1–7 v1.0.1. AtmosphericIQ LLC / Engineering Director, Inc.
Data Legend
Values are displayed with colors from lowest (left) to highest (right)
Interactive Environmental Data Map
Click anywhere on the map to get data values for that location
Location Analysis
Technical Specifications

Global Ground Movement Susceptibility Raster (GMSR) Class 1–7 v1.0.2

Raster File

GMSR_v1_Global_Ground_Movement_Susceptibility_Class_7.tif

Public Dataset URL

https://secure.engineeringdirector.com/public/map/GMSR_v1_Global_Ground_Movement_Susceptibility_Class_7


Product Overview

The Global Ground Movement Susceptibility Raster (GMSR) is a global geospatial dataset representing relative susceptibility to long-term ground movement hazards.

The framework integrates five independently modeled geohazard processes:

  • Landslide Susceptibility
  • Soil Creep Potential
  • Shrink–Swell Potential
  • Subsidence Susceptibility
  • Seismic Amplification

The resulting raster provides a classified susceptibility framework ranging from Class 1 through Class 7, where higher class values indicate greater susceptibility to geotechnical ground movement processes.

The Class 1–7 product provides a simplified categorical representation of the underlying GMSR continuous susceptibility framework for visualization, communication, screening, and portfolio-level risk assessment.

The dataset is intended for:

  • Infrastructure planning
  • Site screening
  • Climate resilience analysis
  • Engineering due diligence
  • Asset portfolio risk assessment
  • Environmental planning
  • Geotechnical hazard screening
  • Catastrophe risk modeling

Dataset Information

Attribute Value
Product Name Global Ground Movement Susceptibility Raster (GMSR) Class 1–7
Version 1.0.2
Coverage Global
Coordinate System WGS 84 (EPSG:4326)
Resolution ~1 km
Raster Format GeoTIFF
Cell Type UInt8
Value Range 1–7
Units Susceptibility Class
NoData Value None
Temporal Basis 2020–2024 Climate Normals and Contemporary Geological Datasets
Update Frequency Periodic
Primary Output Ground Movement Susceptibility Class

Product Status

Attribute Value
Status Production Release
Validation Status Independently Validated
Validation Dataset NASA Global Landslide Catalog
Validation Sample Size 11,001 Independent Landslide Events
Validation Metric AUC (Area Under the Curve)
Validation Result 0.893
Publication Year 2026
Current Version 1.0.2

Raster Statistics

Statistic Value
Minimum Class 1
Maximum Class 7
Number of Classes 7

Class Interpretation

Class Interpretation
1 Very Low Susceptibility
2 Low Susceptibility
3 Low–Moderate Susceptibility
4 Moderate Susceptibility
5 Moderate–High Susceptibility
6 High Susceptibility
7 Very High Susceptibility

Classification Methodology

The Class 1–7 raster is derived directly from the validated GMSR v1 continuous susceptibility raster.

Continuous susceptibility scores were grouped into seven ordinal classes representing increasing levels of relative ground movement susceptibility.

The classification framework is intended to improve interpretability and communication while preserving the relative spatial patterns contained within the continuous raster.


Relationship to Continuous Product

The GMSR Class 1–7 raster is derived directly from the validated GMSR v1 Continuous Susceptibility Raster (0–100).

The continuous raster remains the authoritative scoring product for analytical, API, and engineering workflows.

The Class 1–7 raster is intended primarily for visualization, communication, portfolio screening, reporting, and categorical risk assessment.


Visualization Standard

The public GMSR Class 1–7 raster uses a categorical color scheme aligned with increasing susceptibility.

Class Color Theme
1 Blue
2 Light Blue
3 Blue-Green
4 Cream / Tan
5 Orange
6 Orange-Red
7 Red

The visualization standard is intended for interpretation and communication purposes only and does not alter underlying raster values.

Official GMSR v1 Color Table

```text default 255:255:255 nv 0:0:0:0

1 92:128:184 2 146:170:212 3 194:202:188 4 228:222:165 5 236:184:110 6 228:150:86 7 222:64:54 ```

The color table above represents the official GMSR v1 web visualization standard used within the Engineering Director platform and associated public map services.


Class Interpretation RGB
1 Very Low Susceptibility 92:128:184
2 Low Susceptibility 146:170:212
3 Low–Moderate Susceptibility 194:202:188
4 Moderate Susceptibility 228:222:165
5 Moderate–High Susceptibility 236:184:110
6 High Susceptibility 228:150:86
7 Very High Susceptibility 222:64:54

GMSR Framework

The GMSR framework combines five independently developed susceptibility surfaces into a single composite index.

Component Weighting

Component Weight
Landslide Susceptibility 30%
Soil Creep Potential 20%
Shrink–Swell Potential 15%
Subsidence Susceptibility 15%
Seismic Amplification 20%

Each component raster was independently developed, normalized, and validated prior to integration within the composite framework. The final GMSR raster combines these component products using a weighted susceptibility methodology to provide a globally consistent screening-level assessment of ground movement susceptibility.

Composite Equation

text GMSR = 0.30 × Landslide Susceptibility + 0.20 × Soil Creep Potential + 0.15 × Shrink–Swell Potential + 0.15 × Subsidence Susceptibility + 0.20 × Seismic Amplification

Missing Data Handling

The final composite uses available-data weighted normalization.

Where one or more component datasets contain NoData values, available component weights are automatically normalized to preserve the 0–100 scoring framework.

This methodology minimizes artificial gaps while maintaining global coverage.


Methodology Summary

Landslide Susceptibility

Evaluates terrain susceptible to gravitational mass movement using:

  • Slope Gradient
  • Local Relief
  • Climate Wetness
  • Root Zone Soil Moisture
  • Lithologic Susceptibility

Soil Creep Potential

Evaluates slow long-term downslope soil movement using:

  • Slope Gradient
  • Local Relief
  • Climate Wetness Index
  • Root Zone Soil Moisture
  • Freeze–Thaw Cycles
  • Wet–Dry Cycles
  • Lithologic Susceptibility

Shrink–Swell Potential

Evaluates expansive soil behavior using:

  • SoilGrids Clay Content
  • WRB Vertisol Occurrence
  • Soil Wilting Point
  • Wet–Dry Cycles

Subsidence Susceptibility

Evaluates susceptibility to gradual settlement resulting from:

  • Lowland Topographic Position
  • Wetland Occurrence
  • Soil Saturation Potential
  • River Influence
  • Drainage Position
  • Soft Soil Susceptibility
  • Hydrogeologic Conditions

Seismic Amplification

Evaluates susceptibility to amplified ground shaking using:

  • Global Seismic Hazard
  • Lithologic Susceptibility
  • Slope Gradient
  • Local Relief

Validation

The NASA Global Landslide Catalog was not used during model development, calibration, weighting, or raster generation and therefore provides an independent validation dataset.

The GMSR v1 framework was independently evaluated against documented landslide events from the NASA Global Landslide Catalog.

Validation was performed by comparing GMSR susceptibility scores at landslide event locations against randomly distributed global background locations.

Validation statistics reported below are inherited from the parent GMSR v1 continuous susceptibility raster from which the Class 1–7 product is derived.

Validation Dataset

Attribute Value
Validation Dataset NASA Global Landslide Catalog
Original Records 11,059
Valid Landslide Events Evaluated 11,001
Random Background Locations 33,883
Validation Method Event-Based Spatial Comparison
Validation Metric Area Under Curve (AUC)

Validation Results

Metric Value
Landslide Mean GMSR 36.79
Random Mean GMSR 20.61
Landslide Median GMSR 36.72
Random Median GMSR 21.73
Landslide Class 5–7 (%) 85.41
Random Class 5–7 (%) 21.71
Landslide Class 7 (%) 1.37
Random Class 7 (%) 0.10
AUC 0.893

Understanding AUC

AUC (Area Under the Receiver Operating Characteristic Curve) measures the ability of a model to distinguish between documented hazard locations and randomly distributed background locations.

An AUC value of 0.50 indicates no discriminatory power, while a value of 1.00 indicates perfect separation.

The GMSR v1 validation achieved an AUC of 0.893, indicating strong correspondence between elevated susceptibility scores and observed landslide occurrence.

In practical terms, if a documented landslide location and a random global location are selected at random, there is approximately an 89.3% probability that the landslide location will receive a higher GMSR susceptibility score.

Validation Interpretation

Documented landslide occurrences were strongly concentrated within elevated GMSR susceptibility classes relative to randomly distributed global locations.

Approximately 85% of evaluated landslide events occurred within GMSR Classes 5–7, while only 22% of random background locations occurred within those same classes.

The resulting AUC value of 0.893 indicates strong agreement between elevated GMSR susceptibility scores and observed landslide occurrence at global scale.

These results demonstrate that the GMSR framework successfully identifies regions exhibiting increased susceptibility to ground movement processes and provides independent evidence supporting the utility of the dataset for screening-level geohazard assessment.


Component Products


Source Datasets

Validation Datasets

Derived Internal Datasets


Related Products

Primary GMSR Products


Limitations

  • Intended for screening and planning purposes.
  • Not a substitute for site-specific geotechnical investigation.
  • Does not predict timing of failures.
  • Does not represent actual displacement rates.
  • Does not incorporate local engineering controls.
  • Does not include real-time monitoring information.
  • Validation is based on documented landslide occurrence and does not independently validate all contributing ground movement mechanisms.
  • Susceptibility scores represent relative conditions and should not be interpreted as event probability.

Source Citations

GEBCO

GEBCO Compilation Group.

The General Bathymetric Chart of the Oceans (GEBCO).

International Hydrographic Organization (IHO) and Intergovernmental Oceanographic Commission (IOC).

https://www.gebco.net

GLiM

Hartmann, J., & Moosdorf, N. (2012).

The Global Lithological Map Database (GLiM): A representation of rock properties at the Earth surface.

Geochemistry, Geophysics, Geosystems, 13(12).

https://doi.org/10.1029/2012GC004370

SoilGrids

Poggio, L., de Sousa, L.M., Batjes, N.H., Heuvelink, G.B.M., Kempen, B., Ribeiro, E., & Rossiter, D. (2021).

SoilGrids 2.0: Producing soil information for the globe with quantified spatial uncertainty.

SOIL, 7, 217–240.

https://doi.org/10.5194/soil-7-217-2021

FLDAS

McNally, A., Arsenault, K., Kumar, S., et al. (2017).

A land data assimilation system for food and water security applications.

Scientific Data, 4, 170012.

https://doi.org/10.1038/sdata.2017.12

WRB

IUSS Working Group WRB.

World Reference Base for Soil Resources.

Food and Agriculture Organization of the United Nations (FAO).

https://www.fao.org/soils-portal/data-hub/soil-classification/world-reference-base

GEM

Global Earthquake Model Foundation (GEM).

Global Seismic Hazard Model (GSHM) and Peak Ground Acceleration Products.

Global Earthquake Model Foundation.

https://www.globalquakemodel.org

NASA Global Landslide Catalog

Kirschbaum, D.B., Stanley, T., & Zhou, Y.

The NASA Global Landslide Catalog: A global database of rainfall-triggered landslide events.

National Aeronautics and Space Administration (NASA), Goddard Space Flight Center.

https://data.nasa.gov/dataset/global-landslide-catalog-export

Freeze–Thaw Cycles

Mazzella, J., Mazzella, N. (2026).

Freeze–Thaw Cycles Raster (2020–2024).

AtmosphericIQ LLC / Engineering Director, Inc.

Derived from global climate reanalysis and temperature climatology datasets.

Wet–Dry Cycles

Mazzella, J., Mazzella, N. (2026).

Wet–Dry Cycles Raster (2020–2024).

AtmosphericIQ LLC / Engineering Director, Inc.

Derived from global soil moisture climatology and hydrologic variability datasets.


Version History

Version Date Description
1.0 2026 Initial public release
1.0.1 2026 Added NASA Global Landslide Catalog validation, validation statistics, AUC documentation, finalized public URLs, official visualization standard, and publication-ready metadata.
1.0.2 2026 Updated component methodology documentation following production code audit. Added WRB source dataset and refined framework documentation to reflect validated component model inputs and weighting structures.

Attribution

Joseph Mazzella

Nicole Mazzella

AtmosphericIQ LLC

Engineering Director, Inc.


Dataset Citation

Mazzella, J., Mazzella, N. (2026).

Global Ground Movement Susceptibility Raster (GMSR) Class 1–7 v1.0.2.

AtmosphericIQ LLC / Engineering Director, Inc.


Version Information

Property Value
Dataset Name Global Ground Movement Susceptibility Raster (GMSR)
Dataset Version 1.0.2
Publication Year 2026
Authors Joseph Mazzella, Nicole Mazzella
Organization AtmosphericIQ LLC / Engineering Director, Inc.
Resolution ~1 km
Coordinate System WGS 84 (EPSG:4326)
Coverage Global
Range 1–7
Raster Type Classified
Cell Type UInt8
Primary Output Ground Movement Susceptibility Class
Validation Dataset NASA Global Landslide Catalog
Validation Sample Size 11,001 Independent Landslide Events
Validation AUC 0.893
Processing Framework GMSR v1 Composite Geohazard Framework

Data Distribution Analysis

These histograms show the distribution of pixel values across the entire raster dataset, helping you understand the range and frequency of different measurements.

Linear Scale Distribution
Shows the actual frequency distribution of values using a standard linear scale.
Logarithmic Scale Distribution
Shows the same data using a logarithmic scale, making it easier to see patterns in data with large value ranges.