BIG Terrain Data by Morten Revsbæk, Co-founder and CEO, SCALGO
1 like1,466 views
The document discusses the importance and challenges of analyzing big terrain data, which has transitioned from traditional 30-100 meter data to accurate meter or sub-meter data through technologies like lidar. It highlights the significant value of detailed terrain data for applications such as flood risk analysis, emphasizing the limitations of existing analysis tools in handling large datasets. Additionally, it introduces Scalgo's innovative flash flood mapping product that effectively uses this data for risk assessments, particularly in Denmark.
BIG Terrain Data by Morten Revsbæk, Co-founder and CEO, SCALGO
- 1. BIG Terrain Data Morten Revsbæk Co-founder and CEO June 2013
- 2. 2/9 Turning BIG Terrain Data into Knowledge Detailed terrain Data BIG Analysis Difficult/impossible on national or global scale Knowledge End-user value Analysis
- 3. 3/9 Detailed Terrain Data Available Previously 30-100 meter data Now accurate meter or sub-meter data (e.g. LiDAR) Examples Europe: Denmark, Sweden, Netherlands, … USA: NC, OH, PA, DE, IA, LA, … World: 12m in 2014 by Astrium (EADS) BIG Denmark at 1-meter data: 42 billion data points World at 12-meter data: ~1000 billion data points Availability accelerating Free-data initiatives (Denmark, Iceland, Finland, US states, …) Data continuously improved/updated (e.g. Netherlands)
- 4. 4/9 Detailed Terrain Data Essential Recent USGS (Dewberry) report Billion dollar annual benefit of national US LiDAR Flood risk analysis top benefit area Example: Sea-level rise effect analysis 90-meter data 2-meter data
- 5. 5/9 Big Terrain Data Difficult to Analyze Many analysis tools available (e.g flood risk analysis) Most tools cannot analyze BIG terrain datasets Handles millions of points vs. billions of points in new datasets Typical workarounds Tiling: Break terrain into pieces processed individually Leads to cumbersome workflows Often not possible (e.g. water flow) → Standard cloud processing inadequate → Very little analysis (knowledge extraction) on national/global scale!
- 6. 6/9 Founded by world-leading algorithm researchers MADALGO (Aarhus University, Denmark) World-leading massive data research center Include researchers at MIT and MPI Duke University (USA) Geometric algorithms experts Unique (I/O-efficient) technology Billions of data points (even on desktops) Large-area (national/global) analysis Frequent and cost efficient analysis update SCALGO
- 7. 7/9 Product Example: Flash Flood Mapping Large-area screening of flood risk from extreme rain events Analyses how water gathers in terrain depressions Determines amount of rain before any given terrain part flooded Event based: E.g estimate flooding from 10, 50 or 100 year event Distributed as data service or through SCALGO online Red: 10 year Yellow: 50 year Blue: 100 year
- 8. 8/9 Product Example: Flash Flood Mapping Successfully introduced on Danish market Computed on 2-meter Denmark data (26 billion points) Hundreds of points in family home lot! Cost of 2011 Copenhagen extreme rain event over $1 billion → 1/3 of local government now use SCALGO Flash Flood Mapping Independently verified: accurate, detailed and cost efficient Working to build international market
- 9. 9/9 Turning BIG Terrain Data into Knowledge Detailed terrain Data Essential in analysis Increasingly available BIG Analysis Difficult/impossible on national or global scale Knowledge Valuable e.g. flood risk Data collection driver Analysis
- 10. 10/9 Flash Flood Mapping After 10mm rainAfter 50mm rainAfter 100mm rainAfter 150mm rain
- 11. 11/9 Flash Flood Map with Drainage After 50mm rain