University of Pennsylvania, CIS 565: GPU Programming and Architecture, Final Project
By Xiaoxiao Zou, Keyu Lu, and Jason Li
- Clone this repo
- Run
npm i - Start with
npm start - View the page in your browser!
Camera Controls
Left Mouse: change camera orientationWASD: move camera positionSpace/Shift: move the camera up/downresetCamera: on the GUI, reset the camera to the scene's default position
Simulation Controls
simulate: start/stop simulationSelect Terrain: allow the selection between different terrains to simulate snow accumulationdeltaTime: the time step to use in the simulationtemperature: change the temperature to be used for weather simulationprecipitation: change the precipitation to be used for weather simulationuse GUI weather: enable the use of weather parameters given in the GUI for weather simulationHour of Day, Day of Year: enable the use of change the day of the year which will affect the amount of solar radiation to the snow.Height Multiplier: enable the use of change the height multiplier that changes the height of the terrain.Grid size: change the grid size of each small cell grid size for terrain mesh.Fog Start,Fog End: change the range of the distanced fog to enhance the reality of rendered terrain and snow scene
This project implements a real-time snow accumulation simulation based on the paper: Real-time GIS-based snow cover approximation and rendering for large terrains. The paper uses Unreal Engine as a basis for its simulation, but this project attempts to bring it to the mobile platform of the everyday browser using WebGPU as its framework for both simulation and rendering.
We use compute shaders written in WGSL for WebGPU to simulate our snow accumulation. The overall pipeline can be broken down into 5 stages:
- Terrain Grid Generation: from a .geotiff file with height data, generate a mesh with a uniform grid.
- Compute: Snow Fall: Using simulated precipitation data, a snow-water equivalent (SWE) is calculated per cell
- Compute: Snow Melt: The degree-day method is used to simulate the melting of snow independently for each terrain cell.
- Compute: Snow Redistribution: For each terrain cell, redistribute accumulated snow to neighboring cells using inclination data.
- Terrain & Snow Rendering: Output a color and displacement map from the compute pipeline to be rendered over the terrain.
- Vertex Stage: Instance each mesh grid cell to the corresponding pixel of heightmap and calculated normals, uvs and positions of the terrain mesh.
- Fragment Stage: Using the normals and uvs from vertex stage together with texture from compute pipeline to calculate the final coloring of the snow mountain.
To ensure the efficiency of the snow accumulation simulation, a uniform grid over the terrain is used for computation, and snow values are calculated independently using a compute pipeline for each grid cell.
We create a terrain mesh using data from uploaded .tiff files, which is real-data geotiff in DEM (Digital Elevation Method) format, and use this mesh to create uniform terrain cells over the terrain for snow simulation. Physically-based parameters for each terrain cell such as inclination and aspect are then pre-processed and calculated for use further down in the pipeline for our accumulation simulation. A matching Ariel Imaginary file which has the same real-data geo info as the .tiff file is also used to render terrain.
Using physically-based variables such as inclination, aspect, temperature, and precipitation data, our compute pipeline runs the three stages of snow accumulation described above: snowfall, snow melt, and snow redistribution.
For each terrain cell, a snow-water equivalent (SWE), which represents the amount of water contained in the snowpack per unit area, is calculated, and this value represents the amount of snow output from our compute pipeline. This SWE value is modified through physically based parameters such as weather data for the snowfall stage and solar radiation for the snow melt stage. After these two stages, we then calculate the redistribution of snow between cells using terrain inclination and curvature for our final SWE output. This final step approximates physical behavior such as snow depletion on mountain tops and snow accumulation in gullies. The compute stage of our pipeline outputs a displacement and color map to be used in our snow rendering.
Using our displacement and color map output from the compute shader, we interpolate between the texture of the terrain given from our GeoTIFF data and our snow data. A displacement map is used to simulate the appearance of snow rising and disappearing throughout the seasons of our simulation and can be used to visualize the differences in snow accumulation between different types of terrain. In addition, a skybox is rendered behind our terrain to enhance the visual appearance and atmosphere of our scene.
To further enhance the realisticness of our simulation for snow accumulation on large-scale terrain, we added a separate rendering pipeline for the cube-map-based skybox and integrated it in main.ts to allow the skybox to work with our central pipeline for terrain and snow rendering. We also added distanced fog in the fragment shader and added GUI parameters for the user to control the range of the fog for better interactivity.
We instanced mesh on vertex shader using instancing drawing. The mesh size is limited to 8kx8k on GPU mesh generation (this limit is caused by WebGPU texture maximum size is limited to 8kx8k).
The performance of our program depends on how large the resolution of the height map is.
| Resolution | 265x265 | 4057x3571 | 2918 x 5841 |
|---|---|---|---|
| FPS | 60 | 27 | 20 |
| Memory (MB) | 60 | 390 | 500 |
Tested on Nvidia RTX 4060
- Higher-resolution snow simulation
- Importing of .tiff files and weather data
- Merging of all pipeline components