1999Unpublished venueRequires access

TRMM-based merged precipitation analyses

Robert F. Adler, George J. Huffman, David T. Bolvin, Eric Nelkin, Scott Curtis

Open publisher page 0 citations

Abstract

This paper describes results of using Tropical Rainfall Measuring Mission (TRMM) information as the key calibration tool in a merged analysis on a 1X1 latitude/longitude monthly scale based on multiple satellite sources and raingauge analyses. The TRMM-based product is compared with surface-based validation data sets and the community-based 20-year Global Precipitation Climatology Project (GPCP)monthly analyses. The TRMM-based merged analysis uses the TRMM information to calibrate the estimates from SSM/I and geosynchronous IR observations and merges those estimates together with the TRMM and gauge information to produce accurate rainfall estimates with the increased sampling provided by the combined satellite information. This TRMM merged analysis uses the combined instrument (Precipitation Radar [PR] and TRMM Microwave Imager [TMI]) retrieval of Haddad as the TRMM estimate with which to calibrate the other satellite estimates. This TRMM Combined instrument (TCI) estimate is shown to produce very similar absolute values to the other main TRMM products. The TRMM and other satellites merged analysis compares favorably to the atoll data set of Morrissey for the months of 1998 with a very small positive bias of 2%. However, comparison with the preliminary results from the TRMM ground validation radar information at Kwajalein atoll in the western Pacific Ocean shows a 26% positive bias. Therefore, absolute magnitudes from TRMM and/or the ground validation need to be treated with care at this point. A month by month comparison of the TRMM merged analysis and the GPCP analysis indicates very similar patterns, but with subtle differences in magnitude. Focusing on the Pacific Ocean ITCZ one can see the TRMM-based estimates having higher peak values and lower values in the ITCZ periphery. These attributes also show up in the statistics, where GPCP>TRMM at low values (below 10 mm/d) and TRMM>GPCP at high values (greater than 15 mm/d). Integrated over the 37N-37S belt for all of 1998 the TRMM value is 3.1 mm/day compared to 2.9 mm/day for GPCP, with the results when confined to ocean being 3.3 and 3.0, respectively. Therefore, TRMM has approximately 10% higher tropical oceanic precipitation than GPCP. Examples of TRMM monthly fields in analysis of geographic and temporal variations will also be discussed. Results for 1998 indicate that TRMM has a different ratio of western Pacific to eastern Pacific rainfall than GPCP, perhaps indicating that TRMM's ability to discern the vertical structure of the rain may lead to better estimates in the eastern Pacific. In addition the rainfall anomalies associated with the '98-'99 ENSO event are shown to be larger with the TRMM results as compared to the GPCP analyses.

About this research paper

What this paper is about

This paper describes results of using Tropical Rainfall Measuring Mission (TRMM) information as the key calibration tool in a merged analysis on a 1X1 latitude/longitude monthly scale based on multiple satellite sources and raingauge analyses. The TRMM-based product is compared with surface-based validation data sets and the community-based 20-year Global Precipitation Climatology Project (GPCP)monthly analyses. The TRMM-based merged analysis uses the TRMM information to calibrate the estimates from SSM/I and geosynchronous IR observations and merges those estimates together with the TRMM and gauge information to produce accurate rainfall estimates with the increased sampling provided by the combined satellite information. This TRMM merged analysis uses the combined instrument (Precipitation Radar [PR] and TRMM Microwave Imager [TMI]) retrieval of Haddad as the TRMM estimate with which to calibrate the other satellite estimates. This TRMM Combined instrument (TCI) estimate is shown to produce very similar absolute values to the other main TRMM products. The TRMM and other satellites merged analysis compares favorably to the atoll data set of Morrissey for the months of 1998 with a very small positive bias of 2%. However, comparison with the preliminary results from the TRMM ground validation radar information at Kwajalein atoll in the western Pacific Ocean shows a 26% positive bias. Therefore, absolute magnitudes from TRMM and/or the ground validation need to be treated with care at this point. A month by month comparison of the TRMM merged analysis and the GPCP analysis indicates very similar patterns, but with subtle differences in magnitude. Focusing on the Pacific Ocean ITCZ one can see the TRMM-based estimates having higher peak values and lower values in the ITCZ periphery. These attributes also show up in the statistics, where GPCP>TRMM at low values (below 10 mm/d) and TRMM>GPCP at high values (greater than 15 mm/d). Integrated over the 37N-37S belt for all of 1998 the TRMM value is 3.1 mm/day compared to 2.9 mm/day for GPCP, with the results when confined to ocean being 3.3 and 3.0, respectively. Therefore, TRMM has approximately 10% higher tropical oceanic precipitation than GPCP. Examples of TRMM monthly fields in analysis of geographic and temporal variations will also be discussed. Results for 1998 indicate that TRMM has a different ratio of western Pacific to eastern Pacific rainfall than GPCP, perhaps indicating that TRMM's ability to discern the vertical structure of the rain may lead to better estimates in the eastern Pacific. In addition the rainfall anomalies associated with the '98-'99 ENSO event are shown to be larger with the TRMM results as compared to the GPCP analyses.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

This paper describes results of using Tropical Rainfall Measuring Mission (TRMM) information as the key calibration tool in a merged analysis on a 1X1 latitude/longitude monthly scale based on multiple satellite sources and raingauge analyses. The TRMM-based product is compared with surface-based validation data sets and the community-based 20-year Global Precipitation Climatology Project (GPCP)monthly analyses. The TRMM-based merged analysis uses the TRMM information to calibrate the estimates from SSM/I and geosynchronous IR observations and merges those estimates together with the TRMM and gauge information to produce accurate rainfall estimates with the increased sampling provided by the combined satellite information. This TRMM merged analysis uses the combined instrument (Precipitation Radar [PR] and TRMM Microwave Imager [TMI]) retrieval of Haddad as the TRMM estimate with which to calibrate the other satellite estimates. This TRMM Combined instrument (TCI) estimate is shown to produce very similar absolute values to the other main TRMM products. The TRMM and other satellites merged analysis compares favorably to the atoll data set of Morrissey for the months of 1998 with a very small positive bias of 2%. However, comparison with the preliminary results from the TRMM ground validation radar information at Kwajalein atoll in the western Pacific Ocean shows a 26% positive bias. Therefore, absolute magnitudes from TRMM and/or the ground validation need to be treated with care at this point. A month by month comparison of the TRMM merged analysis and the GPCP analysis indicates very similar patterns, but with subtle differences in magnitude. Focusing on the Pacific Ocean ITCZ one can see the TRMM-based estimates having higher peak values and lower values in the ITCZ periphery. These attributes also show up in the statistics, where GPCP>TRMM at low values (below 10 mm/d) and TRMM>GPCP at high values (greater than 15 mm/d). Integrated over the 37N-37S belt for all of 1998 the TRMM value is 3.1 mm/day compared to 2.9 mm/day for GPCP, with the results when confined to ocean being 3.3 and 3.0, respectively. Therefore, TRMM has approximately 10% higher tropical oceanic precipitation than GPCP. Examples of TRMM monthly fields in analysis of geographic and temporal variations will also be discussed. Results for 1998 indicate that TRMM has a different ratio of western Pacific to eastern Pacific rainfall than GPCP, perhaps indicating that TRMM's ability to discern the vertical structure of the rain may lead to better estimates in the eastern Pacific. In addition the rainfall anomalies associated with the '98-'99 ENSO event are shown to be larger with the TRMM results as compared to the GPCP analyses.

Key concepts: Environmental science, Satellite, Meteorology, Remote sensing, Calibration, Longitude, Precipitation, Sampling (signal processing)

Related papers

Back to paper searchBrowse research topicsOriginal source
TRMM-based merged precipitation analyses — Research Paper | ScholarLens