Systems and methods for generating flood hazard estimation using machine learning model and elevation data

Patent metadata

Field	Value
Publication number	US-2025191258-A1
Application number	US-202519061626-A
Country	US
Kind code	A1
Filing date	Feb 24, 2025
Priority date	Oct 5, 2020
Publication date	Jun 12, 2025
Grant date	—

Abstract

A system and method for flood hazard estimation inputs a satellite elevation map and applies a machine learning model to output a geographic map representing flood hazard areas. The machine learning model is trained to produce an output of a deterministic hazard mapping algorithm. The method retrieves a DEM topography file representing elevation data of an identified terrain, and applies a sink-filling algorithm to detect and fill sinks in the DEM topography. The algorithm subtracts the DEM elevation data to generate a filled topography, and identifies flattest regions of the filled topography. The algorithm then generates a flood hazard map by merging the filled topography and the DEM elevation data, using a weighting function that balances the detected sinks and the flattest regions of the filled topography.

First claim

What is claimed is:

1 . A computer-implemented method comprising:
inputting, by a computer, a first elevation map comprising a plurality of sinks having one or more outlets or spill-points; generating, by the computer, a reference map using a deterministic hazard mapping algorithm; training, by the computer, a machine learning model to transform an input image comprising the first elevation map to an output image of a map representing flood hazard areas using the reference map; and generating, by the computer, the map representing the flood hazard areas based on an input of a second elevation map.

2 . The method of claim 1 , wherein the machine learning model comprises a generative adversarial network to abstract the second elevation map to a flooding-related feature map.

3 . The method of claim 2 , wherein the machine learning model comprises a plurality of skip connections between layers of a deep learning convolutional neural network.

4 . The method of claim 2 , wherein an objective function of the generative adversarial network includes a reconstruction loss with a weighting mechanism that increases importance of high hazard areas.

5 . The method of claim 4 , wherein:
the reconstruction loss comprises an L1 loss function, reweighted to increase the importance of the high hazard areas; and the L1 loss function is reweighted via a weighting factor of w=2.5y+2.5, clamped between 0.02 and 1.

6 . The method of claim 1 , wherein the map representing flood hazard areas identifies geographic areas most vulnerable to pluvial flooding.

7 . The method of claim 1 , wherein the deterministic hazard mapping algorithm comprises:
detecting sinks of the first elevation map; filling the sinks of the first elevation map to a level of their lowest outlet or spill-point; subtracting elevation data of the first elevation map to generate a filled topography file; identifying the flattest regions of the filled topography file; and generating a flood hazard map file using a weighting function to merge the filled topography file and the elevation data, wherein the weighting function balances the detected sinks and the flattest regions of the filled topography file.

8 . The method of claim 1 , wherein:
the machine learning model comprises a deep learning neural network, wherein the first and second elevation maps are digital elevation model (DEM) topography representations of elevation data of an identified terrain, wherein the deep learning neural network applies the deterministic hazard mapping algorithm to the DEM topography to detect and fill sinks in the DEM topography; the deep learning neural network further applies the deterministic hazard mapping algorithm to subtract the elevation data of the DEM topography to generate a hidden layer representing a filled topography map and to identify flattest regions of the hidden layer representing the filled topography map; the deep learning neural network further applies the deterministic hazard mapping algorithm in an output layer of the deep learning neural network merging the hidden layer representing the filled topography map and the DEM topography representing the elevation data of the identified terrain; and merging the hidden layer representing the filled topography map and the DEM topography generates a weighted combination of the filled topography map and the DEM topography balancing the identified flattest regions of the filled topography map versus the detected sinks in the DEM topography.

9 . A computer-implemented method comprising:
detecting, by a computer, sinks of a topography map, the sinks having one or more outlets or spill-points; filling, by the computer, the sinks of the topography maps to a level of their lowest outlet or spill-point; subtracting, by the computer, elevation data of the topography map to generate a filled topography file; identifying by the computer, using a plurality of slope values determined from the topography map, a flat region of the filled topography file; and generating, by the computer, a flood hazard map file using a weighting function to merge the filled topography file and the elevation data, wherein the weighting function balances the detected sinks in DEM elevation data and the flattest regions of the filled topography file.

10 . The method of claim 9 , wherein the flat region of the filled topography file is identified by comparing the plurality of slope values to a predefined threshold.

11 . The method of claim 10 , wherein the predefined threshold does not exceed two percent.

12 . A non-transitory computer-readable medium containing instructions that when executed by one or more processors, causes the one or more processors to perform operations comprising:
inputting a first elevation map comprising a plurality of sinks having one or more outlets or spill-points; generating a reference map using a deterministic hazard mapping algorithm; training a machine learning model to transform an input image comprising the first elevation map to an output image of a map representing flood hazard areas using the reference maps; and generating the map representing the flood hazard areas based on an input of a second elevation map.

13 . The non-transitory computer-readable medium of claim 12 , wherein the deterministic hazard mapping algorithm comprises:
detecting sinks of the first elevation map; filling the sinks of the first elevation map to a level of their lowest outlet or spill-point; subtracting elevation data of the first elevation map to generate a filled topography file; identifying the flattest regions of the filled topography file; and generating a flood hazard map file using a weighting function to merge the filled topography file and the elevation data, wherein the weighting function balances the detected sinks and the flattest regions of the filled topography file.

14 . The non-transitory computer-readable medium of claim 12 , wherein the machine learning model comprises a generative adversarial network to abstract the first elevation map to a flooding-related feature map.

15 . The non-transitory computer-readable medium of claim 14 , wherein the generative adversarial network includes an objective function, wherein in training the machine learning model a generator of the generative adversarial network tries to minimize the objective function against an adversarial of the generative adversarial network that tries to maximize the objective function.

16 . The non-transitory computer-readable medium of claim 15 , wherein the objective function of the generative adversarial network includes a reconstruction loss with a weighting mechanism that increases importance of high hazard areas, the reconstruction loss comprising an LI loss function, reweighted to increase the importance of the high hazard areas.

17 . The non-transitory computer-readable medium of claim 14 , wherein the generative adversarial network comprises a deep learning neural network, wherein the first or second elevation maps are a digital elevation model (DEM) topography representing elevation data of an identified terrain, wherein the deep learning neural network applies the deterministic hazard mapping algorithm to the DEM topography to detect and fill sinks in the DEM topography.

18 . The non-transitory computer-readable medium of claim 17 , wherein the deep learning neural network further applies the deterministic hazard mapping algorithm to subtract the elevation data of the DEM topography to generate a hidden layer representing a filled

Assignees

• Bank Of Montreal

Inventors

• Kenyon-Dean Kian
• Zhao Bo
• Kasiri Keyvan
• Vahlis Yevgeniy
• Fraser Todd
• Morrison Lyndsay
• Torrance Michael
• Wu Stella

Classifications

• G06N3/094

  Adversarial learning · CPC title

• G06N3/0475

  Generative networks · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/0455

  Auto-encoder networks; Encoder-decoder networks · CPC title