During an 11-day mission in 2000 the Space Shuttle Endeavour used a special radar system to obtain detailed near world-wide elevation data. The so-called Shuttle Radar Topography Mission (SRTM) provided a 3 arc-seconds (90-meter at equator) resolution raster terrain dataset from 60° north to 56° south. The raster contains 60.6 billion cells of which 20.8 billion contain real data. It is stored in roughly 14,000 files (tiles) and take up more than 200 gigabytes on disk.
Using SCALGO Hydrology on the SRTM terrain model we assessed the risk of flooding due to rising sea-level by computing for each cell the minimal sea-level rise that results in flooding of the cell (i.e. the level where the cell is connected to the ocean by cells below the level).
We also used SCALGO Hydrology to assess the flooding vulnerability due to extreme rain. We first flooded the terrain, that is, removed (filled) all terrain depressions (in a process that simulates uniformly pouring water on the terrain until a steady-state is reached and all additional added water flow to the boundary of the terrain). We then computed a flow direction for each cell (i.e. the direction water flows from the cell), and finally the flow accumulation for each cell (that is, the amount of water that reaches the cell if a unit of water is initially placed on each cell). Note that assigning flow directions is non-trivial on flat areas, and that the flow accumulation of a cell corresponds to the number of "upstream" cells (i.e. an area). River networks are often extracted as the cells with a flow accumulation value above a certain threshold.
The above computations were performed on the entire SRTM 60.6 billion cell raster dataset without thinning (that is, on the full 3 arc-second resolution dataset) and without use of tiling (where the dataset is broken up into smaller pieces that are processed individually). On a standard workstation with 4GB of main memory and low-cost 7200RPM disk drives, the sea-level rise and the flooding computation of the terrain were performed in approximately half a day, while the flow routing and flow accumulation computation was performed in approximately one day. Discover how you can assess the risk of flooding due to rising sea-level and extreme rain using SCALGO Live's global flood mapping simulation.
Map of the Bay Area in California, USA.
All of the computation steps (sea-level rise resulting in flooding, and flooding of a terrain followed by flow direction and flow accumulation computations) are quite standard, but SCALGO software can perform them on a massive dataset such as SRTM. In the large and impressive HydroSHEDS project of the USGS the later steps are performed on (a modified version of) the SRTM data, but only after thinning the data to 15-arc seconds (500-meter resolution at the equator). One reason the steps are hard to perform on large datasets is that they are "non-local" in the sense that a small local change in topography can affect the result of the computation far away. Thus it is hard to use tiling to perform them, however, exactly due to the "non-local" characteristics of the steps it may be vital to utilize the full resolution. View more Success Stories or find out more about our SCALGO Hydrology package.NEXT: Modeling and Hydrological Analysis of Denmark BACK: Success Stories