This Level-2B swath product (AE_Rain) contains instantaneous measurements of rain rate and rain type (convective vs. stratiform), generated from Level-2A brightness temperatures (AE_L2A).
We kindly request that you cite the use of this data set in a publication using the following citation. For more information, see our Use and Copyright Web page.
Adler, R., T. Wilheit, Jr., C. Kummerow, and R. Ferraro. 2004. AMSR-E/Aqua L2B Global Swath Rain Rate/Type GSFC Profiling Algorithm. Version 2. [indicate subset used]. Boulder, Colorado USA: NASA National Snow and Ice Data Center Distributed Active Archive Center. doi: 10.5067/AMSR-E/AE_RAIN.002.
|Spatial coverage and resolution||This data set offers coverage of all ice-free and snow-free land and ocean between 70°N and 70°S at a resolution of 5.4 km.|
|Temporal coverage and resolution||Temporal coverage is from 18 June 2002 to 3 October 2011. Each swath spans approximately 50 minutes.
For a summary of temporal coverage for various AMSR-E products and algorithms, see the AMSR-E Data Versions Web page.
|Tools for accessing data||For tools that work with AMSR-E data, see the Tools for AMSR-E Data Web page. For general tools that work with HDF-EOS data, see the NSIDC HDF-EOS Web page.|
|Grid type and size||These are swath data.|
|File naming convention||AMSR_E_L2_Rain_X##_yyyymmddhh_f.hdf|
|File size||Each half-orbit granule is approximately 18 MB.|
Rain Type (convective or stratiform)
Reverb | ECHO
Dr. Robert Adler
Mesoscale Atmospheric Processes Branch
Laboratory for Atmospheres
NASA/Goddard Space Flight Center
Greenbelt, MD, USA
2207 Computer and Space Sciences Building
University of Maryland
College Park, MD USA
Dr. Christian Kummerow
Department of Atmospheric Science
Colorado State University
Fort Collins, CO, USA
Dr. Thomas Wilheit, Jr.
Department of Atmospheric Sciences
Texas A&M University
College Station, TX, USA
NSIDC User Services
National Snow and Ice Data Center
CIRES, 449 UCB
University of Colorado
Boulder, CO 80309-0449 USA
phone: +1 303.492.6199
fax: +1 303.492.2468
form: Contact NSIDC User Services
Data are in Hierarchical Data Format - Earth Observing System (HDF-EOS) format. Table 1 uses the following notations to describe the data types:
|Float64:||64-bit (8-byte) floating-point integer|
|Float32:||32-bit (4-byte) floating-point integer|
|Int8:||8-bit (1-byte) signed integer|
|Int16:||16-bit (2-byte) signed integer|
|Time||Float64||approximately 2000 scans||Scan start time in International Atomic Time in seconds with 01 January 1993 00:00:00 as the zero base (TAI93).||n/a||n/a|
|Latitude||Float32||392 x approximately 2000 scans||Latitude (-70.0 to 70.0)||n/a||99|
|Longitude||Float32||392 x approximately 2000 scans||Longitude (-180.0 to 180.0)||n/a||999|
|Rain Rate||Int16||392 x approximately 2000 scans||Rain rate (mm/hr).||Multiply data values by 0.1 to obtain values in mm/hr.||-9999|
|Rain Type||Int8||392 x approximately 2000 scans||Convective rain percentage (0-100).||Multiply data values by 0.01 to obtain values in percent.||-99|
|Rain Status||Int16||392 x approximately 2000 scans||61-67 = Ambiguous rain retrievals for Coast
51 = Ambiguous rain retrievals for Ocean
13-14 = Ambiguous rain retrievals for Land
0 = Rain retrieval is possible
-31 = Mask due to ice/cloud surface
-41 = Mask due to polarization
-51 = Not raining
-61 = Rain retrieval is not possible
|Surface Type||Int8||392 x approximately 2000 scans||
3: sea ice
Note: Beginning with the algorithm version B04, data dimensions changed from 486 pixels to 392 pixels due to the loss of the 89 GHz A-horn.
This section explains the file naming convention used for this product with an example. The date and time correspond to the first scan of the granule.
Example file name: AMSR_E_L2_Rain_V09_200705172328_A.hdf
Refer to Table 2 for the valid values for the file name variables listed above.
|Product Maturity Code (Refer to Table 3 for valid values.)|
|file version number|
|hour, listed in UTC time, of first scan in the file|
|minute, listed in UTC time, of first scan in the file|
|orbit direction flag (A = ascending, D = descending)|
Product Maturity Code
|Preliminary - refers to non-standard, near-real-time data available from NSIDC. These data are only available for a limited time until the corresponding standard product is ingested at NSIDC.|
|Beta - indicates a developing algorithm with updates anticipated.|
|Transitional - period between beta and validated where the product is past the beta stage, but not quite ready for validation. This is where the algorithm matures and stabilizes.|
|Validated - products are upgraded to Validated once the algorithm is verified by the algorithm team and validated by the validation teams. Validated products have an associated validation stage. Refer to Table 4 for a description of the stages.|
|Product accuracy is estimated using a small number of independent measurements obtained from selected locations, time periods, and ground-truth/field program efforts.|
|Product accuracy is assessed over a widely distributed set of locations and time periods via several ground-truth and validation efforts.|
|Product accuracy is assessed, and the uncertainties in the product are well-established via independent measurements made in a systematic and statistically robust way that represents global conditions.|
Table 5 provides examples of file name extensions for related files that further describe or supplement data files.
|Extensions for Related Files||Description|
|.qa||Quality assurance information|
|.ph||Product history data|
Each half-orbit granule is approximately 18 MB.
The following map shows a typical day of coverage with 28 half-orbits.
This data set offers coverage of all ice-free and snow-free land and ocean between 70 degrees north and south latitudes.
Data are 5.4 km resolution.
Temporal coverage is from 18 June 2002 to 3 October 2011. Each swath spans approximately 50 minutes. See the AMSR-E Data Versions Web page for a summary of temporal coverage for different AMSR-E products and algorithms.
Each swath spans approximately 50 minutes. The data sampling interval is 2.6 msec per sample for the 6.9 GHz to 36.5 GHz channels, and 1.3 msec for the 89.0 GHz channel. A full scan takes approximately 1.5 seconds. AMSR-E collects 243 data points per scan for the 6.9 GHz to 36.5 GHz channels, and 486 data points for the 89.0 GHz channel.
The number of satellite passes per day is a function of latitude as shown on the AMSR-E Observation Times Web page.
Satellite-based estimates of rain rate and rain type rely primarily on modeling the absorption and emission effects on microwave signals for specified cloud temperatures, water vapor, and hydro meteor profiles. Atmospheric transmittance windows below 20 GHz, from 30 GHz to 40 GHz, and at 90 GHz are used for rainfall monitoring. Below 20 GHz, rainfall absorption and emission are predominant, and ocean surfaces are warmer than the background radiation. Above 60 GHz, evidence of rainfall is primarily from scattering, where areas of heavy rainfall are colder than their backgrounds. Between 20-60 GHz, a combination of absorption and scattering is present.
A radiative transfer equation that includes absorption and scattering coefficients is the basis for deriving rain rate from brightness temperatures in this data set. The absorption and scattering coefficients, which are summarized in more detail in (Kummerow and Ferraro 2007 ), are expressed as an integral over the range of rain drop sizes. Radiative transfer calculations are used to determine brightness temperatures given atmospheric temperature, water vapor, and hydro meteor profiles. These computations are carried out for the AMSR-E frequencies of 6.9, 10.7, 18.7, 36.6 and 89.0 GHz and 54 degree incidence angle, and for different freezing levels.
At all channels, brightness temperatures increase toward a maximum and then drop off as rainfall rates increase further. The main difference between channels is the range of rainfall rates for which the curve increases in the emission region and decreases in the scattering region (Kummerow and Ferraro 2007 ). The brightness temperature at low frequencies is primarily a function of absorption. The rain rate follows from the absorption coefficient implied by the measurements. Ice and snow are efficient scatterers of microwave radiation compared with rain. Since land background has a high emissivity, rainfall rate over land must be inferred from the ice-scattering signature, instead of relying on the emission signal from rain drops.
Please refer to the AMSR-E Instrument Description document.
The AMSR-E Level 2 rainfall algorithm in rooted in a Bayesian retrieval scheme over oceans and a regression of scattering signals to surface rainfall over land (Wilheit, Kummerow, and Ferraro 2003) . The ocean retrieval relies primarily on the emission signal from the rain drops themselves while the land retrieval relies solely on the scattering of high frequency (89 GHz) radiation from precipitation sized-ice particles at and above the freezing level. The two approaches are needed due to the vastly different surface emissivities and the resulting differences in the sensor information content over ocean and land, respectively. In order to build a consistent algorithm framework the land portion of the algorithm was converted from its original regression form (Grody 1991) and updated by (Ferraro 1997) to a Bayesian framework, but in such a fashion as to reproduce the original regression equations.
Both the land and ocean algorithms begin with a set of Cloud Resolving Model simulations that prescribe the surface rainfall and the associated hydro meteor profiles. Both schemes use texture information to classify observed scenes as either convective, stratiform or mixed. The Bayesian scheme is then invoked to match the observed brightness temperature with entries in the database that match the observed brightness temperature. The difference between ocean and land thus consists chiefly of two aspects. Over land, the possible profiles are narrowed significantly to only those that match the historically derived brightness temperature to rainfall relations, and over land, more empirical relations are needed in order to discriminate raining scenes from brightness temperature depressions caused by radio metrically cold surfaces. The algorithm is discussed in detail in Kummerow and Ferraro (2007) and Wilheit, Kummerow, and Ferraro (2003).
The algorithm over oceans uses a representative set of pre computed Cloud Resolving Model (CRM) profiles to establish the relationship between cloud micro physical parameters and up welling brightness temperatures. Once the database of representative profiles and Brightness Temperatures (Tb) are generated, the algorithm uses a Bayesian inversion methodology in the following manner:
Pr(R|Tb) = Pr(R) * Pr(Tb | R)
Pr(R) = probability that a profile R will be observed
Pr(Tb | R) = probability of observing the set of brightness temperatures given a rain profile R.
The retrieval algorithm thus generates a new cloud profile from the weighted sum of structures in the cloud resolving model data base that are consistent with the observed brightness temperature as well as brightness temperature horizontal variability. The latter is used to determine the convective/stratiform nature of the precipitation. The algorithm is discussed in detail in Kummerow and Ferraro (2007) and Wilheit, Kummerow, and Ferraro (2003)
Accurate rainfall retrievals over land are far more difficult than oceanic retrievals due to the large and variable emissivity of the land surface. Specifically, the high emissivity masks the emission signature that is related directly to the water content in the atmosphere. Instead, only the brightness temperature depression due to scattering in the upper portion of clouds is observable. The scattering increases with increasing frequencies. Consequently, brightness temperature depressions at the 89-GHz channel of AMSR-E contain the least ambiguous signal of scattering by ice and/or large raindrops.
A further complication that arises over land is the lack of consistent backgrounds against which to compare the brightness temperature depression. To alleviate the problem caused by the varying emissivity associated with changes in surface characteristics such as surface wetness, snow cover, vegetation, etc., a rain/no-rain temperature depression threshold is required to screen out false identification of rain. Additionally, snow and desert surfaces cause depressed brightness temperatures at high frequencies due to surface volume scattering, and can be confused with the rain signature. If these surface types are not properly screened, they can be misinterpreted as ice scattering in clouds.
The AMSR-E precipitation team decided to use the same GPROF retrieval methodology as used for the ocean retrieval (Wilheit, Kummerow, and Ferraro 2003). However, unlike the ocean component, the initial database of possible profiles was carefully selected to include only those profiles that fit the empirical relation developed by Ferraro and Marks (1995). Thirty-six profiles, out of the several thousand profiles in the GPROF database, were found to satisfy this relationship (McCollum et al. 1999). The team then computed the expected AMSR-E brightness temperatures for these profiles for use in the a-priori look-up table used in the Bayesian inversion algorithm shown above.
See the AMSR-E Data Versions Web page for a summary of algorithm changes since the start of mission.
Quantifying errors in this data set is complicated, because it involves understanding the nature of precipitation. Uncertainties arise when the rain layer thickness is not well understood, or when inhomogeneous rainfall occurs below the resolution of the satellite. Another potential source of error is the non-precipitating component of clouds, which contribute to brightness temperatures. Scattering-based retrievals over land also present many uncertainties, most notably the lack of a consistent relationship between frozen rain aloft and liquid at lower altitudes. Quantifying the scattering by ice is especially problematic. Ambiguities occur in the data because microwave radiation is scattered not only by rainfall and associated ice, but by snow cover and dry land (Kummerow and Ferraro 2007 ).
A know error exists related to sun glint that results in missing Rain Rate values, and presents as gray ovals in the AE_Rain browse images. Sun glint is not a problem over land; however, the algorithm is using only geometry to determine sun glint causing missing values to exist over land. Sun glint is included in the algorithm because it affects the brightness temperatures out to the missing radius; however, a thorough investigation is yet to be completed. Preliminary investigations indicate that the bias could be up to 15 percent in the affected areas.
Each HDF-EOS file contains core metadata with Quality Assessment (QA) metadata flags that are set by the Science Investigator-led Processing System (SIPS) at the Global Hydrology and Climate Center (GHCC) prior to delivery to NSIDC. A separate metadata file with a .xml file extension is also delivered to NSIDC with the HDF-EOS file, and it contains the same information as the core metadata. Three levels of QA are conducted with the AMSR-E Level 2 and 3 products: automatic, operational, and science. If a product does not fail QA, it is ready to be used for higher-level processing, browse generation, active science QA, archive, and distribution. If a granule fails QA, SIPS does not send the granule to NSIDC until it is reprocessed. Level-3 products that fail QA are never delivered to NSIDC (Conway 2002).
Brightness temperatures are verified to be within the physical bounds (50 K 305 K) for all channels used by the rainfall algorithm. Automated QA for the rainfall algorithm is difficult because heavy rainfall can mask the surface thereby hindering geo-location verification. The rainfall algorithm therefore relies on the Level 2A product for QA of the geo-location. As a final QA check on the computed rainfall, the brightness temperatures of the computed rainfall are compared to the observed brightness temperatures. If the difference between computed and observed brightness temperatures exceeds a pre-defined threshold, the rainfall is set to missing.
AMSR-E Level-2A data arriving at GHCC are subject to operational QA prior to processing higher-level products. Operational QA varies by product, but it typically checks for the following criteria in a given file (Conway 2002):
AMSR-E Level-2A data arriving at GHCC are also subject to science QA prior to processing higher-level products. If less than 50 percent of a granule's data are good, the science Q/A flag is marked suspect when the granule is delivered to NSIDC. In the SIPS environment, the science QA includes checking the maximum and minimum variable values, and percent of missing data and out-of-bounds data per variable value. At the Science Computing Facility (SCF), also at GHCC, science QA involves reviewing the operational QA files, generating browse images, and performing the following additional automated QA procedures (Conway 2002):
Geo-location errors are corrected during Level-2A processing to prevent processing anomalies such as extended execution times and large percentages of out-of-bounds data in the products derived from Level-2A data.
The Team Lead SIPS (TLSIPS) developed tools for use at SIPS and SCF for inspecting the data granules. These tools generate a QA browse image in Portable Network Graphics (PNG) format and a QA summary report in text format for each data granule. Each browse file shows Level-2A and Level-2B data. These are forwarded from the Remote Sensing Systems (RSS) to the GHCC along with associated granule information, where they are converted to HDF raster images prior to delivery to NSIDC. The QA summary reports are available on the GHCC AMSR-E Web page.
Please refer to AMSR-E Validation Data for information about data used to check the accuracy and precision of AMSR-E observations.
Kummerow, C., Y. Hong, W. S. Olson, S. Yang, R. F. Adler, J. McCollum, R. Ferraro, G. Petty, D. B. Shin, and T. T. Wilheit. 2001. The Evolution of the Goddard Profiling Algorithm (GPROF) for Rainfall Estimation from Passive Microwave Sensors. Journal of Applied Meteorology 40: 1801-1820.
Kummerow, C., W. Olson, and L. Giglio. 1996. The Evolution of the Goddard Profiling Algorithm (GPROF) for Rainfall Estimation from Passive Microwave Sensors. IEEE Transactions on Geosciences and Remote Sensing 34: 1213-1232.
Kummerow, C., and R. Ferraro. 2006. [Supplement] EOS/AMSR-E Level-2 Rainfall: Algorithm Theoretical Basis Document. Fort Collins, Colorado, USA: Colorado State University. (PDF file, 245 KB)
For more information regarding related publications, see the Research Using AMSR-E Data Web page.
The following acronyms are used in this document:
|AMSR-E||Advanced Microwave Scanning Radiometer - Earth Observing System|
|CRM||Cloud Resolving Model|
|EOS||Earth Observing System|
|FTP||File Transfer Protocol|
|GHCC||Global Hydrology and Climate Center|
|GSFC||Goddard Space Flight Center|
|HDF-EOS||Hierarchical Data Format - EOS|
|NASA||National Aeronautics and Space Administration|
|NSIDC||National Snow and Ice Data Center|
|PNG||Portable Network Graphics|
|RSS||Remote Sensing Systems|
|SCF||Science Computing Facility|
|SIPS||Science Investigator-led Processing System|
|TLSIPS||Team Lead SIPS|
|UTC||Universal Time, Coordinated|