1. Introduction
Over past decades, significant progress has been made in satellite precipitation product development (Joyce and Arkin 1997; Ferraro et al. 1998; Iguchi et al. 2000; Kummerow et al. 2001). Temporal resolution and timely availability have been improved by blended techniques (e.g., Xie and Arkin 1997; Hsu et al. 1999; Kidd et al. 2003; Joyce et al. 2004; Huffman et al. 2007; Marzano et al. 2007; Turk and Mehta 2007). The resulting products, such as near-real-time precipitation products, are widely used in various research and applications (Hsu et al. 1999; Sorooshian et al. 2000; Kidd et al. 2003; Joyce et al. 2004; Huffman et al. 2007; Turk and Mehta 2007). However, the lack of support for user-defined areas or points of interest poses a major obstacle to quickly gaining knowledge of product quality and behavior on a local or regional scale.
The National Aeronautics and Space Administration (NASA) Goddard Earth Science Data and Information Services Center (GES DISC) is home to the Tropical Rainfall Measuring Mission (TRMM) product archive (Vicente et al. 2007). GES DISC provides not only TRMM data, but also value-added data services to its users (Liu et al. 2007) such as the TRMM Online Visualization and Analysis System (TOVAS; http://disc2.nascom.nasa.gov/Giovanni/tovas/). Many users compare their own gauge or radar data with TRMM products before using them in their research or applications. At present, users have to download original data and software, which can be time consuming, costly, and especially difficult for users from developing countries with limited resources.
Existing rainfall validation and intercomparison online services (e.g., http://www.bom.gov.au/bmrc/SatRainVal/validation-intercomparison.html) focus primarily on providing pregenerated maps, plots, and statistics for a quick and easy view of large areas. With an Internet (“Web”) browser, users can quickly learn how satellite precipitation products are compared with gauge data on a daily basis, without traditional data downloading and processing. Because nearly all satellite products are grouped together in one place and can be updated daily, users find it convenient to examine how each product performs by clicking on links associated with maps, plots, and statistics. Some sites also provide monthly and seasonal verifications for those who want to see product performance over a longer time period.
However, existing Web services usually cover very large areas, such as Australia, the United States, western Europe, South America, and so on, and do not adequately address similar issues at a local or regional scale. Environmental conditions, such as land surface types, topography, and precipitation regimes, vary from region to region. Further, existing static Web services do not provide interactive and customized services. Is there an easier way for users to evaluate precipitation products and accelerate the process from research to applications?
This paper describes ongoing work at GES DISC to develop an online information system prototype for the validation and intercomparison of global satellite precipitation algorithms. This prototype provides users with customized information on the expected bias and accuracy of the products and gives algorithm developers a better understanding of the strengths and weaknesses of different algorithmic approaches and data sources. Section 2 of this paper describes the system and data. Section 3 presents an example, and section 4 discusses conclusions and future plans.
2. System description and data
To address NASA earth science challenges, mission science teams and other data users need easy access to data from NASA earth science missions. In support of this goal, GES DISC has begun developing a family of reliable, low-cost, Web-based tools that are easy to use and powerful in their capabilities. These tools have already significantly increased the productivity of users of water-cycle-related NASA data. For example, TOVAS provides global rainfall data and information ranging from historical to near–real time to users around the world (Zhang et al. 2005; Huffman et al. 2007; Liu et al. 2007; Yin et al. 2008; Meier and Knippertz 2009).
The Online Precipitation Intercomparison Tool (OPIT; available at http://disc2.nascom.nasa.gov/Giovanni/tovas/rain.ipwg.shtml) is a main component of the online information system prototype for the validation and intercomparison of global satellite precipitation algorithms (Fig. 1). Other components provide data archiving, processing, and scheduling. The OPIT system architecture is based on TOVAS (Liu et al. 2007), which primarily consists of TRMM products. TOVAS has been in operation since March of 2000.
TOVAS is part of the GES DISC Interactive Online Visualization and Analysis Infrastructure (“Giovanni”; http://giovanni.gsfc.nasa.gov) (Acker and Leptoukh 2007; Liu et al. 2007; Berrick et al. 2009). The principal design goal for Giovanni was to provide a quick and simple interactive means for science data users to study various phenomena by trying various combinations of parameters measured by different instruments, arrive at a conclusion, and then generate graphs suitable for publication (e.g., Mills et al. 2004; Peck and Congdon 2005; additional Giovanni-related peer-reviewed papers are available online at http://disc.sci.gsfc.nasa.gov/giovanni/additional/publications). In addition, Giovanni provides a way to ask relevant “what if” questions and to get answers that stimulate further investigation, all without having to download and preprocess large amounts of data.
Figure 2 shows a schematic system diagram of OPIT, consisting of Hypertext Markup Language (HTML) and Common Gateway Interface (CGI) scripts written in Perl and Grid Analysis and Display System (GrADS) scripts (http://grads.iges.org/grads/). OPIT includes an image-map Java applet through which users select a bounding-box area to define an area of interest (Fig. 1).
GrADS was chosen for its widespread use as a tool that provides easy access, manipulation, and visualization of meteorological observation and model data. It supports a variety of data formats such as binary, GRIB, NetCDF, HDF, and HDF-EOS.
By using the OPIT Web interface, a user selects one or more datasets, the spatial area, the temporal extent, and the type of output. Table 1 lists the supported functions and data output type. The selection criteria are passed to the CGI scripts for processing.
OPIT data (http://www.bom.gov.au/bmrc/SatRainVal/IPWG_precip_archive.html) are provided by the International Precipitation Working Group (IPWG) (http://www.isac.cnr.it/~ipwg/). Currently available products in OPIT are daily multisatellite rainfall (TMPA-RT, version 5; global, 2005), daily rain gauge (North America, 2005), and daily radar rain (North America, 2005). TMPA-RT is a 3-hourly near-real-time TRMM Multisatellite Precipitation Analysis product (Huffman et al. 2007) from which the daily accumulation is derived. The daily rain gauge product is a Cressman (1959) analysis of daily rain gauge data over the United States and Mexico used as the reference for validating rainfall over the United States (Higgins et al. 2000; ftp://cics.umd.edu/pub/DATA/Validation/documentation/gauge.doc). The daily radar rain product is a composite of “Stage II” Next Generation Weather Radar (NEXRAD) over the United States. Stage II includes large-scale bias corrections from gauge data (http://wwwt.emc.ncep.noaa.gov/mmb/ylin/pcpanl/QandA/).
3. Case study
The 2005 Atlantic hurricane season was the most active hurricane season on record. Hurricane Katrina, which directly hit New Orleans, Louisiana, on 29 August 2005, was the costliest hurricane in U.S. history, despite the fact that it had been downgraded to a category-3 storm by the time it made landfall. OPIT enables examination of how the near-real-time TMPA-RT performs relative to gauge and radar products in the Katrina landfall area.
Figures 3a–c show the daily rainfall totals of TMPA-RT, gauge, and radar products, respectively, on 29 August 2005, revealing large discrepancies among the three products. The radar product indicates the highest amount of rainfall, followed by TMPA-RT and the gauge, which the two-product differencing function confirms. In Fig. 4a, the TMPA-RT rainfall in the hurricane area exceeds that of the gauge product by as much as 90 mm. The radar rainfall exceeds that of TMPA-RT only in the hurricane landfall area but is lower in the surrounding areas (Fig. 4b). Koschmieder (1934), World Meteorological Organization (1962), Wilson (1954), Allerup and Madsen (1979), and Dunn and Miller (1960) have examined underestimation errors of rain gauges as a function of wind speed and found that the error can be highly variable; for example, the error can be 50% (Wilson 1954) or 100% (Allerup and Madsen 1979) for wind speeds ranging from 15 to 25 m s−1.
Figure 5 shows the intercomparison between TMPA-RT and the gauge product in August 2005 at a location near New Orleans (30°N, 90.5°W). Note the Katrina landfall event on 29 August. The time series, the difference plot, and the scatterplot (Figs. 5a–c, respectively) show that the TMPA-RT rainfall greatly exceeds that of the gauge product during the heavy-rain events at the beginning of the month and during the Katrina event. However, during the lightest-rainfall events, the gauge product rainfall exceeds that of TMPA-RT. By contrast, the intercomparison between TMPA-RT and the radar product presents a complicated scenario as shown in Fig. 6. In general, TMPA-RT agrees with the radar product very well through most of the month (Fig. 6a). However, both behave differently during heavy-rainfall events. The TMPA-RT rainfall exceeds that of the radar product at the beginning of the month but is lower during the Katrina event (Figs. 6b,c, respectively), which is consistent with Fig. 4b.
4. Conclusions and future plans
A prototype of the Online Precipitation Information Tool for global satellite precipitation algorithm validation and intercomparison has been developed. Despite its limited functionality and datasets, users can generate customized plots within the United States for 2005. Users can download customized data in ASCII format for further analysis, such as comparison with gauge data. Additional IPWG precipitation data products and updates of current products will be added in the next development phase. IPWG community verification algorithms for satellite precipitation products (Ebert 2007) are not currently included in OPIT; however, the current system architecture has demonstrated the capability to integrate algorithms into OPIT. The authors plan to collaborate with precipitation algorithm scientists and the IPWG to include such algorithms.
A further goal is permitting users to upload their observational data, such as time series of observations (at a point or averaged over a region) or two-dimensional data arrays in the same resolution in ASCII format. This will allow individuals to conduct their own intercomparison tasks without downloading and processing original satellite precipitation data.
OPIT can be expanded to intercompare other near-real-time satellite precipitation products, such as hourly and 3-hourly products, that are widely used in near-real-time monitoring activities and hydrological models. Knowledge of product biases could improve understanding of derived products, such as those from hydrological models (Nijssen and Lettenmaier 2004).
Biases or discrepancies in algorithms and applications need to be identified and quantified. This will require the analysis of additional ancillary data and intermediate products from which the final precipitation is derived, such as environmental conditions, land surface types, and so on. As an initial step, data archived at GES DISC need to be made available within OPIT. For example, the global 4-km and half-hourly merged IR (brightness temperature) dataset is an intermediate product for TMPA-RT (Janowiak et al. 2001); although users can access these data online (http://disc.sci.gsfc.nasa.gov/hurricane/trmm_quikscat_analysis.shtml), integration to OPIT is still needed to allow easier, seamless access.
Another planned feature is the ability to intercompare old and new versions of data products when an algorithm team provides updates. This service will enable users to understand how changes in the data products, such as biases, offsets, and changing spatial/temporal distributions, impact their specific application or research focus.
GIS support for precipitation data is increasingly important as the spatial resolution of satellites improves; detailed geographic information will be beneficial to applications and decision-making activities.
Detailed system and product documentation is essential for users to understand how OPIT works, and it will be made available and transparent to users. Improving OPIT’s service to the user community will require ongoing collaboration with users to implement their feedback and suggestions, which will guide future decisions about additional improvements in OPIT.
Acknowledgments
The authors thank IPWG for assistance in providing GrADS-ready data archives for OPIT. Thanks also are extended to Steven Lloyd and Diane Bruton for improving the manuscript and to two anonymous reviewers for their constructive comments and suggestions. The NASA GES DISC supported this work.
REFERENCES
Acker, J. G., and G. Leptoukh, 2007: Online analysis enhances use of NASA earth science data. Eos, Trans. Amer. Geophys. Union, 88 , 14. doi:10.1029/2007EO020003.
Allerup, P., and H. Madsen, 1979: Accuracy of point precipitation measurements. Danish Meteorological Institute Climatology Paper 5, Charlottenlund, Denmark, 84 pp. [Available from Danish Meteorological Institute, Lyngbyvej 100, DK 2100 Copenhagen, Denmark].
Berrick, S. W., G. Leptoukh, J. D. Farley, and H. Rui, 2009: Giovanni: A Web service workflow-based data visualization and analysis system. IEEE Trans. Geosci. Remote Sens., 47 (1) 106–113.
Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev., 87 , 367–374.
Dunn, G. E., and B. I. Miller, 1960: Atlantic Hurricanes. Louisiana State University Press, 326 pp.
Ebert, E. E., 2007: Methods for verifying satellite precipitation estimates. Measuring Precipitation from Space—EURAINSAT and the Future, V. Levizzani, P. Bauer, and F. Joseph Turk, Eds., Advances in Global Change Research Series, Vol. 28, Springer, 345–356.
Ferraro, R. R., E. A. Smith, W. Berg, and G. J. Huffman, 1998: A screening methodology for passive microwave precipitation retrieval algorithms. J. Atmos. Sci., 55 , 1583–1600.
Higgins, R. W., W. Shi, E. Yarosh, and R. Joyce, 2000: Improved United States precipitation quality control system and analysis. NCEP/Climate Prediction Center Atlas, NCEP, No. 7, 40 pp.
Hsu, K., H. V. Gupta, X. Gao, and S. Sorooshian, 1999: Estimation of physical variables from multichannel remotely sensed imagery using a neural network: Application to rainfall estimation. Water Resour. Res., 35 , 1605–1618.
Huffman, G. J., and Coauthors, 2007: The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8 , 38–55.
Iguchi, T., T. Kozu, R. Meneghini, J. Awaka, and K. Okamoto, 2000: Rain-profiling algorithm for the TRMM precipitation radar. J. Appl. Meteor., 39 , 2038–2052.
Janowiak, J. E., R. J. Joyce, and Y. Yarosh, 2001: A real-time global half-hourly pixel-resolution infrared dataset and its applications. Bull. Amer. Meteor. Soc., 82 , 205–217.
Joyce, R., and P. A. Arkin, 1997: Improved estimates of tropical and subtropical precipitation using the GOES precipitation index. J. Atmos. Oceanic Technol., 14 , 997–1011.
Joyce, R., J. E. Janowiak, P. A. Arkin, and P. Xie, 2004: CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high spatial and temporal resolution. J. Hydrometeor., 5 , 487–503.
Kidd, C., D. R. Kniveton, M. C. Todd, and T. J. Bellerby, 2003: Satellite rainfall estimation using combined passive microwave and infrared algorithms. J. Hydrometeor., 4 , 1088–1104.
Koschmieder, H., 1934: Methods and results of definite rain measurements. Mon. Wea. Rev., 62 , 5–7.
Kummerow, C., and Coauthors, 2001: The evolution of the Goddard profiling algorithm (GPROF) for rainfall estimation from passive microwave sensors. J. Appl. Meteor., 40 , 1801–1820.
Liu, Z., H. Rui, W. L. Teng, L. S. Chiu, G. Leptoukh, and G. Vicente, 2007: Online visualization and analysis: A new avenue to use satellite data for weather, climate, and interdisciplinary research and applications. Measuring Precipitation from Space—EURAINSAT and the Future, V. Levizzani, P. Bauer, and F. Joseph Turk, Eds., Advances in Global Change Research Series, Vol. 28, Springer, 549–558.
Marzano, F. S., D. Cimini, and F. J. Turk, 2007: Multivariate probability matching for Microwave Infrared Combined Rainfall Algorithm (MICRA). Measuring Precipitation from Space—EURAINSAT and the Future, V. Levizzani, P. Bauer, and F. Joseph Turk, Eds., Advances in Global Change Research Series, Vol. 28, Springer, 269–279.
Meier, F., and P. Knippertz, 2009: Dynamics and predictability of a heavy dry-season precipitation event over West Africa—Sensitivity experiments with a global model. Mon. Wea. Rev., 137 , 189–206.
Mills, M. M., C. Ridame, M. Davey, J. La Roche, and R. J. Geider, 2004: Iron and phosphorus co-limit nitrogen fixation in the eastern tropical North Atlantic. Nature, 429 , 292–294. Erratum, 435, 232.
Nijssen, B., and D. P. Lettenmaier, 2004: Effect of precipitation sampling error on simulated hydrological fluxes and states: Anticipating the Global Preciptation Measurement satellites. J. Geophys. Res., 109 , D02103. doi:10.1029/2003JD003497.
Peck, D. R., and B. C. Congdon, 2005: Colony-specific foraging behaviour and coordinated divergence of chick development in the wedge-tailed shearwater Puffinus pacificus. Mar. Ecol. Prog. Ser., 299 , 289–296.
Sorooshian, S., K. L. Hsu, X. Gao, H. V. Gupta, B. Imam, and D. Braithwaite, 2000: Evaluation of PERSIANN system satellite-based estimates of tropical rainfall. Bull. Amer. Meteor. Soc., 81 , 2035–2046.
Turk, F. J., and A. V. Mehta, 2007: Toward improvements in short-time scale satellite-derived precipitation estimates using blended satellite techniques. Measuring Precipitation from Space—EURAINSAT and the Future, V. Levizzani, P. Bauer, and F. Joseph Turk, Eds., Advances in Global Change Research Series, Vol. 28, Springer, 281–290.
Vicente, G. A., and GES DAAC Hydrology Data Support Team, 2007: Global satellite datasets: Data availability for scientists and operational users. Measuring Precipitation from Space—EURAINSAT and the Future, V. Levizzani, P. Bauer, and F. Joseph Turk, Eds., Advances in Global Change Research Series, Vol. 28, Springer, 49–58.
Wilson, W. T., 1954: Discussion of paper “Precipitation at Barrow, Alaska, greater than recorded”. Trans. Amer. Geophys. Union, 35 , 203–207.
World Meteorological Organization, 1962: Precipitation measurements at sea. Tech. Note 47, WMO-124-TP-55, 18 pp.
Xie, P., and P. A. Arkin, 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78 , 2539–2558.
Yin, Z-Y., X. Zhang, X. Liu, M. Colella, and X. Chen, 2008: An assessment of the biases of satellite rainfall estimates over the Tibetan Plateau and correction methods based on topographic analysis. J. Hydrometeor., 9 , 301–326.
Zhang, X., M. A. Friedl, C. B. Schaaf, A. H. Strahler, and Z. Liu, 2005: Monitoring the response of vegetation phenology to precipitation in Africa by coupling MODIS and TRMM instruments. J. Geophys. Res., 110 , D12103. doi:10.1029/2004JD005263.
OPIT Web interface. Users can select a region/point of interest, products, plot type, and beginning and ending dates. Users can generate customized plots by adjusting color bar and y axis. ASCII output of data is available for users to conduct additional analysis. A “help” button provides descriptions about each plot type and option. A data description section provides brief information about data and links to download original data. A non-Java interface is available for those who experience difficulties using Java in their browsers.
Citation: Journal of Applied Meteorology and Climatology 48, 12; 10.1175/2009JAMC2244.1
A schematic diagram of OPIT. Through the Web graphical user interface (GUI), users can conduct analysis and interact with OPIT.
Citation: Journal of Applied Meteorology and Climatology 48, 12; 10.1175/2009JAMC2244.1
The lat–lon map function allows users to plot multiple maps in the same output window for easy comparison. The maps show the rainfall (mm day−1) from the different data products (a) TMPA-RT, (b) gauge, and (c) radar during Hurricane Katrina.
Citation: Journal of Applied Meteorology and Climatology 48, 12; 10.1175/2009JAMC2244.1
Differences (mm day−1) in lat–lon maps can reveal spatial differences between two products, e.g., (a) between TMPA-RT and gauge and (b) between TMPA-RT and radar.
Citation: Journal of Applied Meteorology and Climatology 48, 12; 10.1175/2009JAMC2244.1
Three different plot types for TMPA-RT and gauge at a point near New Orleans (30°N, 90.5°W) in August 2005: (a) overlay of time series (mm day−1), (b) difference of time series (mm day−1), and (c) scatterplot. Broken lines represent points with no data available.
Citation: Journal of Applied Meteorology and Climatology 48, 12; 10.1175/2009JAMC2244.1
As in Fig. 5, but showing differences between TMPA-RT and radar.
Citation: Journal of Applied Meteorology and Climatology 48, 12; 10.1175/2009JAMC2244.1
Supported functions and data output capability.
