1. Chair: Peter Bauer Rapporteur: Joerg Schulz Participants: Geoff PegramArthur Hou Carlos AngelisChristian Mitrescu Francisco J TapiadorSteven Miller Ben JouChristian Kummerow Bizzarro BizzarriMichael Goodman Kyle HilburnEric Smith Shannon BrownAmy Doherty Tristan L’EcuyerUna O’Keeffe Research Working Group

2. Issues I.There is an ongoing uncertainty in quantitative precipitation estimation, quantitative precipitation forecast and the monitoring of precipitation trends on a climatological scale. Radars in space have provided the impetus for most recent advances in rainfall remote sensing. To that end this panel recommends the deployment of next generation space radar systems (e.g., dual frequency and increased sensitivity) such as those proposed for GPM and EGPM for improved understanding of precipitation physics and enhancement of existing radiometer measurements. II.At mid to high latitudes snowfall constitutes an important part of the hydrological cycle. The currently available sensors are inadequate for snowfall detection and quantification over land surfaces. The new space-borne radars and radiometers (e.g. sounding frequencies on EGPM) are needed to continue improvement of the understanding of solid precipitation.

3. Topics The working group identified the following topics as crucial for the adequate use of existing and future space-borne observations: I.Open rainfall retrieval algorithm architecture II.Combining MW-IR-models III.Transportable rainfall estimation techniques IV.General Recommendations

4. From the requirement of across-sensor/satellite consistent estimates and internal physical consistency we recommend an open rainfall retrieval algorithm structure. a)To combine the strengths of research groups in the international community. b) We recommend that this structure be used for the monitoring of the rainfall environment and rainfall generation processes. Processes and environment are critical to further improve our knowledge of precipitation physics and to constrain precipitation retrieval algorithms (frozen precipitation, light rain, precipitation in complex terrain, warm rain etc.). I.Open rainfall retrieval algorithm architecture

5. I.Open rainfall retrieval algorithm architecture, cont’d This has large implications on algorithm development and on validation (e.g., GPM supersites) Action: Establish international community working group for the development of an algorithm framework and validation requirements. (example GPM parametric algorithm development effort, link to IPWG). (ST-LT: C. Kummerow, P. Bauer) We recommend the consideration of NWP-type analysis techniques (such as data assimilation) for the generation of rainfall products from rainfall and rainfall related observations as well as ancillary data with different error structures. These methods can also be used for retrieval error estimation and the optimization of future satellite sensor design (e.g., using observing system simulation studies).

6. I.Open rainfall retrieval algorithm architecture, cont’d IPWG 2002, Standards/Common Procedures Recommendation Recommendation 1: Provide a generally accessible platform on data and algorithms for the research community Action: Define elements of algorithm transition required for general applications. (QPE, bias/error estimate, limits of applicability, documentation, comprehensive bibliography, compliance with user requirements as defined by WMO, EUMETSAT, NASA-DAO, ESA, others - compile user requirements w.r.t. to precipitation on the IPWG web site). (ST: Peter Bauer, Bob Adler, IPWG-O)

7. I.Open rainfall retrieval algorithm architecture, cont’d IPWG 2002, Standards/Common Procedures Recommendation Recommendation 2: Enable co-operation and training through software libraries Action: Provision of radiative transfer codes, algorithm codes, footprint matching codes (e.g. to be used with direct read out raw data, especially for developing countries) with corresponding training tools and documentation. (ST-MT: Joerg Schulz, Ralf Bennartz, IPWG-O) Recommendation 3: Develop new strategies on flexible and global structures for physical algorithm development, validation, and data fusion. Action: Establish a framework for: - (physical) algorithm development (global and regional). (ST-LT: GPM, Vincenzo Levizzani) - a testbed for algorithm validation (comparison data sets etc.). (ST-LT: IPWG-V, IPWG-R)

8. I.Open rainfall retrieval algorithm architecture, cont’d IPWG 2002, Research Recommendation Recommendation 1: The principal problem of under-constrained precipitation retrievals requires a different type of analysis/retrieval approach. (TOR: d) Recommendation 2: The physical uncertainty modelling has to be pursued in a similar fashion as physical algorithm development is carried out. (TOR: d) IPWG 2002, Future Sensors Recommendation 1: From the work of the group indications for future sensors should emerge based on the identified scientific outstanding areas (TOR: d, e) Action: Generate a framework for the development of a multi-frequency simulator, orbit configurations, required to optimise future sensors w.r.t. to different application areas. (MT-LT: IPWG-R)

9. II.Combining MW-IR-models We recommend combined approaches that embody the open and modular architecture philosophy (see I.). For example techniques that combine the quantitative precipitation estimate from the microwave sensor and the temporal information on cloud development from the IR sensor embody the above principles. This will facilitate data reprocessing/reanalysis. We further recommend that merging of model data into retrieval algorithms should be considered for certain applications. The impact of merging procedures on statistics of merged products (e.g., consistency of rain pdf´s, rain occurrence between product and ingredients, scale discrepancies) should be considered for all future algorithm developments.

10. II.Combining MW-IR-models, cont’d IPWG 2002, Standards/Common Procedures Recommendation 3: Develop new strategies on flexible and global structures for physical algorithm development, validation, and data fusion. Action: Product merging and blending. (ST-LT: IPWG-O, Bob Adler, Phil Arkin, Joe Turk)

11. III.Transportable rainfall estimation techniques There is an ongoing demand for high-frequency satellite rainfall retrievals (currently IR-based). These IR techniques should ultimately be replaced by microwave and sub-millimetre sounding from geostationary satellites to provide more direct information on precipitation physics. High frequency observations are mainly used for short-range meteorological and hydrological forecasting applications. We recommend the development of a transportable technique that reflects the existing skill that is well documented, tutorial, and simple to implement. Action: As part of the algorithm development group, we will form a sub-group to implement such an algorithm for a test location (e.g., to be identified by D. Hinsman).

12. General recommendations Action: In the short term, we recommend the planning of a workshop dedicated to frozen precipitation physics and observation (link to GPCP activities). (ST: Ralf Bennartz, Ralph Ferraro) Action: The initiation of a liaison with ITWG for the coordination of common projects. (Represent IPWG at the next ITSC in May 2005, ITWG website). (ST- LT:Amy Doherty, Una O’Keeffe) CGMS to IPWG: GPCP assessment is thought to belong to the Operations WG Snowfall, frozen precip (dedicated workshop GPCP/IPWG) Orographic precipitation Ongoing validation is thought to belong to the Validation WG Formats, Level 2 Precipitation File Content Standard, precipitation standards We support the introduction of standards for precipitation products, e.g., formats. We recommend the formation of a working group under the auspices of GPM to realize format definition.

13. General recommendations, cont’d IPWG 2002, Future Sensors Recommendation 1: From the work of the group indications for future sensors should emerge based on the identified scientific outstanding areas (TOR: d, e) Action: Definition and provision of experimental data sets (ST: Bizzarro Bizzarri) Recommendation 2: In view of future sensor development a long-term strategy for frequency protection has to be developed and integrated in the current ITWG activities. (TOR: e) Actions: Expression of need for frequency protection including approximate frequency bands and their scientific justification w.r.t. future precipitation missions. (very ST: IPWG) Demonstrate the usefulness of chosen frequencies in order to support the request for protection. (ST-MT: Bizzarro Bizzarri)

14. IPWG 2002, Standard/common procedures Recommendation 1: Provide a generally accessible platform on data and algorithms for the research community (TOR: a, d) Actions: Define criteria of satellite data (including meta-data) access for research applications including data availability (schedules), sensor specifications, calibration/validation activities, data formats, data accessibility (cost). (ST: Ralph Ferraro, Ralf Bennartz, IPWG-O). In the short term, the responsible agencies have to be approached for providing information on the following sensors which will be used for experimental/operational applications (Jim Purdom, CGMS): SSMIS (DOD/NOAA), AMSR-E (NASA/NASDA), AMSR-J (NASA/NASDA), MSMR, VIS/IR sensors. Actions: Set up an inventory of field campaign data, co-located satellite data. (ST: IPWG-V) General recommendations, cont’d