Advancing Extreme Weather Monitoring from Space: From TRMM to GPM

Dalia Kirschbaum and Arthur Hou
NASA Goddard Space Flight Center, Greenbelt, Maryland

TRMM satellite image of Tropical Cyclone Yasi on Feb. 1-3, 2011

Figure 1: TRMM satellite image of Tropical Cyclone Yasi on Feb. 1-3, 2011 (left to right), as it made landfall over Queensland, Australia. TRMM’s TMI and PR instruments observed Cyclone Yasi as it developed from a Category 3 tropical cyclone (Feb. 1s, left), to a Category 5 event (Feb. 2, middle) when it made landfall with wind gusts reported at up to 186 mph, and then finally as it began to dissipate on Feb. 3 (right). Image credit: Hal Pierce.

I. Precipitation measurement from space

Within the past five years, extreme hydrometerological events have been responsible for more than 254,000 fatalities worldwide, accounting for over 50% of total natural hazard fatalities [1]. As recent as January 2011, the world has witnessed widespread and deadly landslides outside of Rio de Janeiro, Brazil; devastating floods across eastern Australia; and a Category 5 tropical cyclone impacting Queensland, Australia (see Figures 1 and 2). Knowing when, where, and how much it rains or snows is vital to understanding the impacts of these extreme weather events as well as predicting their future behavior with respect to natural and anthropogenic climate signals.

The 2007 Intergovernmental Panel on Climate Change (IPCC) AR4 Synthesis Report states that anticipated regional-scale changes as a result of global warming include a “very likely” increase in the frequency of heavy precipitation and precipitation at high latitudes and a “likely” increase in tropical cyclone intensity with “likely” decreases in precipitation over subtropical land regions [2]. Understanding the current variability of extreme precipitation events is a critical component in better modeling future extreme weather scenarios. Space-borne sensors provide the means for obtaining such quantitative precipitation estimates over the globe on sub-daily to decadal time scales. Remotely-sensed precipitation products are especially valuable in regions with a dearth of rain-gauge or radar networks, where extreme meteorological events frequently contribute to significant social and economic losses.

The Tropical Rainfall Measuring Mission (TRMM), launched in 1997, uses active and passive microwave instruments to confirm the validity of tropical rainfall estimates derived from operational passive microwave sensors and provides a foundation for merging rainfall information from other satellites. Building upon TRMM’s success, the Global Precipitation Measurement (GPM) Mission will offer a new generation of space-borne precipitation measurements using a constellation of satellites in conjunction with an advanced radar-radiometer system on a GPM “Core Satellite” to deliver near real-time estimates of rain and snow every 2-4 hours anywhere on the globe [3]. GPM is designed to advance scientific understanding of the Earth’s water and energy cycle and to improve predictions of weather, climate, and freshwater resources. In this article, we highlight some of the scientific achievements of TRMM’s space-borne radar-radiometer system and discuss how GPM will enhance extreme weather monitoring and prediction.

II. Cyclone Tracking

High-resolution microwave information is important for fixing the locations of tropical cyclone tracks as well as establishing accurate initial positions to effectively forecast tropical cyclone progression. The TRMM Microwave Imager (TMI) is designed to “see” through clouds and provide high-resolution rainfall structure information within tropical cyclones. TMI data are currently used operationally by NOAA’s National Hurricane Center (NHC), the Joint Typhoon Warning Center (JTWC), and tropical cyclone centers in Japan, India, Australia, etc. for detecting the location and intensity of tropical cyclones and for fixing the position of tropical cyclone tracks. As an example, 16% of all tropical cyclone position fixes by the JTWC in 2004 were from TRMM data [4].

 pre- and post-event images (a, b) over Teresópolis, Brazil follwoing landslides

Figure 2: Several communities outside of Rio de Janerio, Brazil, experienced extensive landslides following an extreme rainfall event early on Jan. 12, 2011. Satellite imagery from the Advanced Land Imager (ALI) on NASA’s EO-1 satellite captured these pre- and post-event images (a, b) over Teresópolis, Brazil. The TMPA real-time precipitation product recorded a peak rainfall event in this area within the same time period. Credit: NASA Earth Observatory image, using EO-1 ALI data provided courtesy of the NASA EO-1 team.

TRMM’s Precipitation Radar (PR) is designed to provide vertical structure information within tropical cyclones. PR data are used to verify the cloud and precipitation structure in mesoscale tropical cyclone models as well as to understand the distribution of latent heat release within tropical cyclones. PR data also have been employed to observe deep convective towers in the eyewall, which have been linked to subsequent rapid deepening of storms [5]. Together, TRMM PR and TMI data have been used to establish key characteristics of the distribution and variability of rainfall in tropical cyclones and obtain insights into storm structure, intensity, and environmental conditions.

The GPM Mission will extend tropical cyclone tracking and forecasting capabilities into the middle and high latitudes, enabling tropical cyclone monitoring as cyclones transition into mid-latitude systems. The GPM Core Observatory will fly in a non-Sun-synchronous orbit at a 65° inclination (TRMM currently flies at a 35° inclination). This orbit will enable diurnal sampling of precipitation and provide new insight into how tropical cyclones may intensify or weaken as they transition into mid-latitude systems.

III. Flood and landslide monitoring

Extreme precipitation over hours to months can trigger disastrous flooding and/or landslides depending on the area’s topography, land use, and climate. Several merged, multi-satellite products have been developed to provide near real-time precipitation information valuable for modeling and monitoring these hydrometeorological events. These include TRMM’s Multi-Satellite Precipitation Analysis (TMPA) product [6], CMORPH (CPC MORPHing technique) [7], and GSMaP (Global Satellite Mapping of Precipitation) [8], among others.

Several hydrological models have been developed to represent flooding events using flood routing schemes or precipitation threshold information. A flood algorithm has been developed that uses TMPA data within the hydrologic model CREST to estimate areas of potential flooding at the global scale [9]. Other hydrologic models have used satellite precipitation information at local to regional scales, highlighting the importance of comprehensive rainfall information in accurate hydrologic predictions [10]. Flood warning systems such as the International Flood Network use satellite precipitation threshold information to alert communities to potential flooding at regional and global scales. Regional models such as EUMETSAT’s Satellite Application Facility on Support to Operational Hydrology and Water Management are working to develop several assimilation techniques that merge multiple satellite inputs in order to offer products and tools for understanding extreme hydrological events.

While rainfall-triggered landslides tend to be local-scale features, a recent threshold-based global algorithm has shown that satellite rainfall may provide useful information for near real-time hazard assessment over larger areas [11]. A global landslide hazard forecasting algorithm has been developed to identify locations of potential landslide activity, coupling TMPA rainfall data with a static susceptibility map based on surface information including topography, soil type, and land cover [12]. Algorithm outputs and documentation of the global flood and landslide systems are available at

Evaluation of both regional and global flood and landslide assessment techniques suggests that merged satellite precipitation products have the potential to inform hydrologic models and hazard monitoring. However, in many instances, higher resolution precipitation inputs are needed at either sub-daily time scales or smaller spatial footprints to deliver accurate rainfall intensities and accumulations for improved monitoring and prediction [13]. GPM near real-time products will provide the spatial and temporal resolution as well as the latitudinal coverage to advance these prototype systems and improve error characterization of algorithm outputs.

IV. Numerical Weather Prediction

Numerical weather prediction (NWP) centers around the world, including the European Centre for Medium-Range Weather Forecasts (ECMWF), the Japan Meteorological Agency (JMA), and the National Center for Environmental Prediction (NCEP), are currently assimilating or plan to assimilate microwave-based satellite rainfall information (radiances or retrievals) into their global or regional models to enhance operational weather forecasts. While studies have shown that rainfall assimilation can improve forecasts of extreme weather events such as hurricanes using a variety of techniques (e.g. [14; 15]), continued advancement in assimilation methodology in the coming decade is expected to further evolve the use of precipitation data in NWP. In addition, many NWP agencies are now creating precipitation products for “nowcasting” to meet the needs of a wider user community, including weather forecasters, hydrologists, farmers, the military, numerical modelers, and the climate community [4]. By providing more accurate and frequent observations in near real-time, GPM will enable numerical weather prediction centers to continue improving operational forecasts, including predictions of extreme weather events.

V. Conclusions

On Jan. 12th, 2011, the equivalent of one month of rainfall fell within one day in Teresópolis and Nova Friburgo, Brazil, triggering thousands of landslides and killing roughly 900 people. These types of extreme hydrometerological events are becoming more and more common as heavy precipitation combined with expanding populations and development aggravate already vulnerable areas. TRMM precipitation products have been used within a broad range of communities to improve tropical cyclone tracking and forecasting, advance natural hazard assessment, and inform numerical weather prediction.

Current satellites provide better than 3-hour mean revisit times over 45% of the globe. GPM’s satellite constellation will enable the collection of coincident measurements over a wider range of latitudes, bringing coverage up to 90% during the peak year of the mission. The advanced radar-radiometer measurement system onboard the GPM Core Satellite, scheduled for launch in 2013, will offer greater measurement sensitivity to light rain and falling snow. Through enhanced spatial and temporal precipitation measurements, the GPM mission will provide rapidly disseminated precipitation products to advance predictive capabilities for natural hazards and extreme weather events from the tropics to high latitudes. Additional information about both TRMM and GPM can be accessed at


[1] EM-DAT, “The OFDA/CRED International Disaster Database,” Université Catholique de Louvain, 2011.

[2] IPCC, “Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change,” R. K. Pachauri, and A. Reisinger, Eds. IPCC, Geneva, Switzerland, 2007.

[3] A. Hou, G. Skofronick-Jackson, C.D. Kummerow, and J.M. Shepherd, “Global precipitation measurement,” Precipitation: Advances in Measurement, Estimation, and Prediction, S. Michaelides, Springer-Verlag, 2008, pp. 131-169.

[4] National Research Council, “Assessment of the benefits of extending the Tropical Rainfall Measuring Mission: A perspective from the research and operations communities: Interim Report,” Washington, D.C., 2006, pp. 103.

[5] O.A. Kelly, J. Stout, and J.B. Halverson, “Tall precipitation cells in tropical cyclone eyewalls are associated with tropical cyclone intensification,” Geophysical Research Letters, vol. 31, 2004.

[6] G.J. Huffman, R.F. Adler, D.T. Bolvin, G. Gu, E.J. Nelkin, K.P. Bowman, Y. Hong, E.F. Stocker, and D.B. Wolff, “The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-Global, Multiyear, Combined-Sensor Precipitation Estimates at Fine Scales,” Journal of Hydrometeorology, vol. 8, 2007, pp. 38-55.

[7] R.J. Joyce, J.E. Janowiak, P.A. Arkin, and P. Xie, “CMORPH: A Method that Produces Global Precipitation Estimates from Passive Microwave and Infrared Data at High Spatial and Temporal Resolution,” Journal of Hydrometeorology, vol. 5, 2004, pp. 487-503.

[8] K. Okamoto, T. Iguchi, N. Takahashi, K. Iwanami, and T. Ushio, “The Global Satellite Mapping of Precipitation (GSMaP) project,” 25th IGARSS Proceedings, 2005, pp. 3414-3416.

[9] Y. Hong, R. Adler, and G.J. Huffman, “Applications of TRMM-based Multi-satellite Precipitation Estimation for Global Runoff Simulation: Prototyping a Global Flood Monitoring System,” Satellite Applications for Surface Hydrology, F. Hossain and M. Gebremichael, Springer Verlag, 2010, p. 327.

[10] F. Hossain and E.N. Anagnostou, “Assessment of current passive-microwave- and infrared-based satellite rainfall remote sensing for flood prediction,” Journal of Geophysical Research, vol. 109, 2004, pp. 1-14.

[11] D.B. Kirschbaum, R. Adler, Y. Hong, and A. Lerner-Lam, “Evaluation of a preliminary satellite-based landslide hazard algorithm using global landslide inventories,” Natural Hazards And Earth System Sciences, vol. 9, 2009, pp. 673-686.

[12] Y. Hong, R. Adler, and G. Huffman, “Evaluation of the potential of NASA multi-satellite precipitation analysis in global landslide hazard assessment,” Geophysical Research Letters, vol. 33, 2006, pp. 1-5.

[13] F. Hossain and D.P. Lettenmaier, “Flood prediction in the future: Recognizing hydrologic issues in anticipation of the Global Precipitation Measurement mission,” Water Resources Research, vol. 42, 2006, pp. 1-5.

[14] A. Hou, S. Zhang, and O. Reale, “Variational continuous assimilation of TMI and SSM/I rain rates: Impact on GEOS-3 hurricane analyses and forecasts,” Monthly Weather Review, vol. 132, 2004, pp. 2094-2109.

[15] P. Bauer, P. Lopez, A. Benedetti, A. Salmond, and E. Moreau, “Implementation of 1D+4D-Var assimilation of precipitation-affected microwave radiances at ECMWF. II: 4D-Var,” Quart. J. Roy. Meteor. Soc., vol. 132, 2006, p. 2307–2332.

Dalia Kirschbaum is the Applications Scientist for the GPM mission. Her research focuses on rainfall-triggered landslide modeling using satellite precipitation data. Arthur Hou is the U.S. Project Scientist for the GPM mission.