1. Development of the Flash Flood Potential (FFPI) The Flash Flood Potential Index for Central NY and Northeast PA by by Jim Brewster WFO Binghamton, NY Moneypenny Creek Flash Flood – May 2004

2. Background Flooding is WFO Binghamton’s #1 High Impact Hazard Flooding is WFO Binghamton’s #1 High Impact Hazard Central NY and Northeast PA have highly variable geography, land cover and use. Central NY and Northeast PA have highly variable geography, land cover and use. Steep, rocky terrain along with flatter sandy plains Steep, rocky terrain along with flatter sandy plains Areas of urbanization Areas of urbanization Wide range of forest cover Wide range of forest cover Similar soil types Similar soil types Experienced forecasters understand some areas are more prone to flash flooding than others. Experienced forecasters understand some areas are more prone to flash flooding than others. But, Where and to what extent? But, Where and to what extent?

3. Flash Flood Potential Index (FFPI) Developed by hydrologist Greg Smith, CBRFC (2003). Developed by hydrologist Greg Smith, CBRFC (2003). He understood that geographical features play a key role in flash flooding. He understood that geographical features play a key role in flash flooding. Developed as background information to be incorporated into production of better gridded Flash Flood Guidance Developed as background information to be incorporated into production of better gridded Flash Flood Guidance Using the FFPI, the roles of land, vegetation and urbanization in flash flooding are visualized. Using the FFPI, the roles of land, vegetation and urbanization in flash flooding are visualized. “Guesswork” to the flash flood problem is reduced “Guesswork” to the flash flood problem is reduced

4. Methodology for creating FFPI Collected available geographic data sets Collected available geographic data sets Using GIS Using GIS Projected datum to Albers Equal Area Projected datum to Albers Equal Area Re-sampled data sets to 90 meter resolution Re-sampled data sets to 90 meter resolution Reclassified data to a standard index Reclassified data to a standard index Mathematically develop a new geographic index grid…the FFPI Mathematically develop a new geographic index grid…the FFPI

5. The Data Four geographic data sets : Four geographic data sets : Slope derived from the USGS DEM (Digital Elevation Model) Slope derived from the USGS DEM (Digital Elevation Model) MLRC Land Use/Land Cover Grid MLRC Land Use/Land Cover Grid AVHRR Forest Density Grid AVHRR Forest Density Grid STATSGO Soil Type Classification STATSGO Soil Type Classification

6. Slope Index Exponentially scaled from 1-10 Exponentially scaled from 1-10 USGS & engineering studies USGS & engineering studies ~30% slope is rated strong-very strong slope. ~30% slope is rated strong-very strong slope. Approx 20 o angle. Approx 20 o angle. Indexed >30% as 10. Indexed >30% as 10.

7. Indexed Land Use/Land Cover Much of region shares a similar index Much of region shares a similar index Mixed forest & grassland. Mixed forest & grassland. Mild-Moderate effect on hydrology Mild-Moderate effect on hydrology Swamp/water 1-2 Swamp/water 1-2 Urban areas 8-10 Urban areas 8-10

8. Indexed Forest Density High density forest areas are given a low potential flood index. High density forest areas are given a low potential flood index. Higher uptake rate of moisture Higher uptake rate of moisture Low density areas are given high potential index. Low density areas are given high potential index. Lower uptake rate of moisture Lower uptake rate of moisture

9. Indexed Soil Texture Class ClassFFPI 1 – Sand2 2 – Loamy Sand4 3 – Sandy Loam3 4 - Silty Loam4 5 – Silt5 6 – Loam6 7 – Sandy Clay Loam7 8 – Silty Clay Loam7 9 – Clay Loam8 10 – Sandy Clay7 11 – Silty Clay8 12 – Clay9 13 – Organic Matter5 14 – Bedrock10

10. Methodology Review Weight average the geographic layers. Weight average the geographic layers. FFPI = (1.5*Slope + LC + Soils + Forest)N FFPI = (1.5*Slope + LC + Soils + Forest)N Local adjustment to calculation Local adjustment to calculation Reviewed against historical events Reviewed against historical events Flash flooding occurs in our forested areas. Flash flooding occurs in our forested areas. Is that element really much of an influence here? Is that element really much of an influence here? FFPI = (1.5*Slope + LC + Soils + 0.5*Forest)/N FFPI = (1.5*Slope + LC + Soils + 0.5*Forest)/N Raw grid is then zonally averaged into the FFMP basins. Raw grid is then zonally averaged into the FFMP basins.

11. 90 Meter Resolution 90 Meter Resolution Warm colors = High Potential Warm colors = High Potential Cool colors = Low Potential Cool colors = Low Potential

12. FFPI mapped to FFMP Basins Fit our conceptual flash flood model. Fit our conceptual flash flood model. New realizations, especially the areas of low FF potential. New realizations, especially the areas of low FF potential. Differentiates the “best of the worst” basins in an area generally known for high flash flood potential. Differentiates the “best of the worst” basins in an area generally known for high flash flood potential.

13. FFPI Versatility  Flexible Formatting  ESRI Shape file  KML/KMZ  GeoTif  Google Earth  BMP  GRx Analyst  Flexible Formatting  ESRI Shape file  KML/KMZ  GeoTif  Google Earth  BMP  GRx Analyst

14. Case Example

15. Fatal Gorge Flood

16. Storm Total Precipitation

18. 90 m High Resolution High resolution GIS imagery of FFPI can be useful to customers and partners for identifying local, potential flood problem areas that may not be already known. High resolution GIS imagery of FFPI can be useful to customers and partners for identifying local, potential flood problem areas that may not be already known.

19. Summary The FFPI was developed in Binghamton due to the important need to have a static geophysical reference grid which better illustrates how local earth system features contribute to flash flooding. The FFPI was developed in Binghamton due to the important need to have a static geophysical reference grid which better illustrates how local earth system features contribute to flash flooding. The FFPI is best used operationally when mapped to the AWIPS FFMP basins for comparison with other flash flood tools and techniques. The FFPI is best used operationally when mapped to the AWIPS FFMP basins for comparison with other flash flood tools and techniques. Through GIS technology, the index could be developed by any local office. Through GIS technology, the index could be developed by any local office. Note: The original developer has a project underway with Office of Hydrology to further develop and expand the FFPI nationally. Note: The original developer has a project underway with Office of Hydrology to further develop and expand the FFPI nationally.

20. Questions ? Reduced false alarms Reduced false alarms Two warnings - Pike County, PA and Oneida County, NY were not issued. Follow-up confirmed no flooding Two warnings - Pike County, PA and Oneida County, NY were not issued. Follow-up confirmed no flooding Increased Lead Time Increased Lead Time Boosted forecaster confidence that additional rain would lead to flash flooding (Warn on Forecast) – Major flash flooding resulted in Delaware County, NY Boosted forecaster confidence that additional rain would lead to flash flooding (Warn on Forecast) – Major flash flooding resulted in Delaware County, NY First Year Performance