121 and 139

Task 121 and 139

Global Floods and Landslides

Principal Investigator(s):

R. Adler, H. Wu


F. Policelli

Last Updated:

October 26, 2012 15:25:28

Description of Problem

The objective is to develop, test and apply a global system for estimating floods and landslides using satellite precipitation information, other satellite data and hydrological models.

Scientific Objectives and Approach

Develop and test hydrological models with satellite precipitation data and evaluate results in terms of floods and landslides with calculations made in real-time. This work is accomplished with colleagues at ESSIC and at Goddard Space Flight Center.


The primary goal of this on-going activity is to adapt, improve, validate, and transfer a global flood and landslide technique to an operational entity so that it will be available for decision-making in support of disaster risk reduction activities around the globe. Regional applications focus on East Africa through our participation in SERVIR-Africa. Our flood and landslide models are routinely running in real-time, with 3-hr updates, with graphic results appearing on the TRMM web site (trmm.gsfc.nasa.gov). Multi-satellite rain estimates are obtained from the TRMM/GPM Precipitation Processing System (PPS) and combined with the surface information (SRTM DEM, MODIS vegetation, etc.) and the models to produce the locations of probable flood and landslide activity.

Results from the project are being used by the International Red Cross and other national and international organizations to monitor global hazard events and plan mitigation strategies. This work is also part of the Societal Benefits efforts under the CEOS/GEO umbrella, with specific items in the international work plan for continuation and testing of the algorithms and transfer of the operational responsibility for the products.

A major accomplishment this year has been the further quantitative evaluation of the real-time global flood monitoring system (GFMS) in terms of flood event detection during the TRMM era (1998-2010) using a global retrospective simulation (3-hourly and 1/8 degree spatial resolution) with the TMPA 3B42V6 rainfall. Four methods were explored to define flood events from the model results, including three percentile-based statistic methods and a Log Pearson-III flood frequency curve method. The evaluation showed the GFMS detection performance improves with longer flood durations and larger affected areas. The impact of dams was detected in the validation statistics. The presence of dams tends to result in more false alarms and false alarm duration. The GFMS statistics for flood durations > 3 days and for areas without dams vary across the four identification methods, but center around a POD of ~ 0.70 and a FAR of ~ 0.65. When both flood events-based categorical verification metrics and flood duration metrics are considered, a method using the 95th percentile runoff depth plus two parameters related to variability and basin size (method 3) may be more suitable for application to our routine, real-time flood calculations. The evaluation showed the GFMS detection performance improves with longer flood durations and larger affected areas. The new GFMS (operationally available at http://trmm.gsfc.nasa.gov/) improved not only the flood detection performance, but also in the presentation of flood evolution (start, development and recession) in the drainage network. The new GFMS is further evaluated with more quantitative flood properties including flood peak timing, peak stage, peak volumes, and duration of flood flows at basins across USA, by comparing the simulated results to gauge based hydrograph. This model development and evaluation provide a pathway forward for continued improvement. The improvements brought by the new hydrological model encourage us to use more physically-based hydrological models to potentially achieve better flood forecasting capability and performance in future endeavor, though very likely with much higher computational cost. The gauge flow based evaluation clearly indicated the importance of snowmelt process in flood simulation which is lack in the CREST model current used in the GFMS.

To realize the potential of global flood monitoring systems, robust flow routing schemes that contain minimal calibration parameters wherever possible are needed, in addition to the a prior parameters. Therefore during this year, to improve the stream flow simulation accuracy, a new kinematic wave based routing scheme for large scale applications while taking into account subgrid stream routing at finer resolution has been developed and tested with a community hydrologic model –Variable Infiltration Capacity model (with snowmelt) using conventional meteorological forcing inputs. Together with the routing scheme developing, an upscaled global hydrography dataset (including flow direction, river network, drainage area, river length, channel slope, overland slope etc.) which are critical inputs for macroscale hydrologic modeling like this study were developed by applying the hierarchical Dominant River Tracing (DRT) algorithm to the HydroSHEDS fine scale (1km) baseline hydrography inputs (Wu et al., 2011).

Refereed Journal Publications

Hong, Y., R.F. Adler, G.J. Huffman, 2009: Applications of TRMM-based Multi-satellite Precipitation Estimation for Global Runoff Simulation: Prototyping a Global Flood Monitoring System, in Satellite Applications for Surface Hydrology, F. Hossain and M. Gebremichael, Eds., Springer Verlag, 327p.

Kirschbaum, D. B., R. Adler, Y. Hong, and A. Lerner-Lam, 2009: Evaluation of a Preliminary Satellite-Based Landslide Hazard Algorithm Using Global Landslide Inventories. Natural Hazards and Earth System Science, 9, 673-686.

Yilmaz, K., R. Adler, Y. Tian, Y. Hong, H. Pierce, 2010: Evaluation of a Satellite-based Global Flood Monitoring System. International Journal of Remote Sensing, accepted.

Wang J., Y. Hong., L. Li., J.J. Gourley., K. Yilmaz., Khan S. I., Policelli. F. S, Adler R. F., Habib S., Irwin. D., Limaye. S. A., Korme.T , and L. Okello, 2011, The Coupled Routing and Excess STorage (CREST) Distributed Hydrological Model. Hydrol. Science Journal,. In press.

Wu, H., R. Adler, Y. Tian, Y. Hong, F. Policelli, 2011a. Evaluation of Satellite-based Real-time Global Flood Detection and Prediction with an Improved Hydrological Model. J. Hydromet. (in review).

Wu, H., Kimball, J. S., Mantua, N., & Stanford, J. (2011). Automated upscaling of river networks for macroscale hydrological modeling. Water Resources Research, 47(3), 1-18. doi:10.1029/2009WR008871