Kick-off Meeting of the project IOMASA, held November 4-5, 2002 at Institute of Environmental Physics (IUP), University of Bremen. Participants: IUP: Dr. Georg Heygster Hong Gang Lars Kaleschke Prof. Klaus Künzi Dr. Christian Melsheimer Dr. Junggang Miao Prof. Justus Notholt Nathalie Selbach DTU-DCRS (Danish Centre for Remote Sensing, Technical University of Denmark): Dr. Leif Toudal DMI (Danish Meteorological Institute): Søren Andersen met.no [DNMI] (The Norwegian Meteorological Institute): Dr. Harald Schyberg SMHI (Swedish Meteorological and Hydrological Institute): Dr. Nils Gustafsson Dr. Tomas Landelius --------------------------------------------------------------------------- AGENDA ======== Introductory items Welcome addresses by coordinating Partner and EU Tour-the-table for presentation of attendees Short reminder of the origin of the initiative Project Overview: Summary and Innovation Project logic graphic Project Structure: 5 Parts and 4 Phases Work breakdown and Deliverables, see table, ~ 1/WP Presentation of the five Project Parts by the respective steering committee members, each presentation covering: - state of the art and partner's experience relevant to IOMASA - progress since submission of proposal - partner's aspiration with regard to IOMASA - short description of the project part, - detailed plan for Phase 1 (WP X.1), including state of planned deliverables Part 1 (IUP): Remote sensing of atmospheric parameters: - Total Water Vapour (TWV) - Cloud Liquid Water (CLW) - Surface emissivity at T-sounder frequencies Part 2 (met.no, SMHI): Improving numerical weather prediction models - TWV assimilation - Temperature Sounding - Surface Flux Part 3 (DTU): Empirical models for emissivity and backscatter of sea ice: - Forward model - Snow - AMSR (together with IUP) Part 4 (DMI + met.no): Sea ice concentration retrieval Part 5 (DTU + DMI): Demonstration of real time processing and user interface. Review of Phase 1: Are we ready to envisage Phase 1, especially Part 5? Internal communication may be required Action items for KOM: Decide on - Period of offline investigation: 2 years, probably 1999 and 2000 - Region of investigation: most parts of the Arctic as covered by the operational model of met.no - Geographic projection grid of all data Social objectives and technical implementation plan Success of project measured in 2 dimensions: - publications in reviewed journals - transfer of scientific progress to applications TIP background, European paradox, role of TIP 3 social objectives: improve (1) weather forecast (2) ice charts (3) heat flux estimation annual versions of TIP required, eTIP User advisory group purpose: dialogue with end users in the a.m. fields will attend PM1, MR, FP potential members: - ECMWF (1) (name?) - HIRLAM initiative (1) (name?) - EUM SAF NWP (1) (name?) - ice service of met.no (2) (name?) - ice service of DMI (2) (name?) - Missing: representatives for (3) (name?) other potential members? Contribution to EU OF Cluster Purpose, participation in workshops, also by other partners 3rd EuroGOOS Conference, Athens, 3-6 Dec (Presentation) Bruxelles meeting 'European Research 2002' (Poster) Questions and answers with the EC Project management: Report duties, table of reports Next meetings: Fix places and dates Any other business End of formal part, optional scientific discussions ------------------------------------------------------------------ 1.1 Welcome Address (K. Künzi). =============================== 1.2 Introductory round. ======================== 1.3 Origin of the Initiative (G. Heygster) ========================================== - Several previous projects on remote sensing of polar regions at IUP, e.g.: PELICON (1996-98), SEA LION (1998-2000) -- both in cooperation with DTU. - Combine remote sensing of polar atmosphere and remote sensing of sea ice and apply it to numerical weather prediction (NWP); IOMASA initiative was actually spawned at a meeting at DMI in July 2001 with participation of DTU, IUP and DMI. 1.4 Overview of IOMASA (G. Heygster) ==================================== - 5 parts, 4 phases, schedule. 2. Presentations about the 5 Project Parts: =========================================== 2.1 Part 1: Remote Sensing of Atmospheric Parameters (G. Heygster) ================================================================== - Part 1 done by IUP - Problem: Sparse data over Arctic region, standard algorithms to derive atmospheric parameters from remote sensing data do not work over ice - Total water vapour (TWV) over ice: New algorithm developed at IUP, derive TWV from ratios of channel differences of SSM/T2 (-> J. Miao et al. 2001) - works up to TWV of 6kg/m^2 (using 150 and 183 GHz channels) - could get to higher values if 89 GHz channel used, but more contribution from surface with unknown emissivity - surface emissivities have been measured during SEPOR/POLEX campaign in March 2001 (-> N. Selbach) - Cloud signature (related to cloud liquid water, CLW) over ice: New algorithm developed at IUP, uses polarisation channel differences and ratios ("R-factor") of SSM/I (J. Miao et al. 2000). - Cloud signature: linear combination of water vapour and small contribution of CLW - Possible improvements: Use AMSR instead of SSM/I (more channels, higher spatial resolution) - Temperature profile retrieval (from SSM/T1) could possibly be improved by surface emissivity information from SSM/I (sea ice) and OLS (surface temperature) and sea ice emissivity knowledge. Questions/Comments: ------------------- (T. Landelius): Wouldn't information on the sea ice emissivity also be useful for AMSU-B humidity sounding? (J. Miao): We only do TWV, not water vapour profiles, but yes. But maybe SSM/I frequencies too far off AMSU-B frequencies? 2.2 Part 2: Improving NWP Models (N. Gustafsson, H. Schyberg) ============================================================== - Part 2 done by met.no and SMHI [state of the art NWP: * models are more sensitive to dynamic fields (motion, temperature) than to static ones; therefore, water vapour until now not yet assimilated * 3D variational assimilation (3D-Var) or 4D-Var (ECMWF, Japan) or * Continuous assimilation (nudging) (DWD) or * flow-dependent assimilation (ensemble Kalman filter) (Canada) for assimilation of satellite and other data] - NWP at met.no and SMHI: HIRLAM (3D-Var operational, 4D-Var soon) - assimilation of data from AMSU-A, ground-based GPS, MODIS TWV - ATOVS data from NOAA-15,16,17 - ATOVS data source: * NOAA/NESDIS (BUFR format), too much time delay: up to 6 hours * via own receiving station (HRPT format), quick, but not global * EUMETSAT retransmission (http://www.eumetsat.de/en/dps/atovs.html), starting Nov.2002, covers North Atlantic and Arctic (to be extended further) - RTTOV-7 model for simulation of ATOVS radiances - surface emissivity needed - Surface flux parameterisation used by met.no and SMHI in HIRLAM: ISBA (Interactions between Soil-Biosphere-Atmosphere) * different soil and vegetation types * snow model * "tiles": various surface types within one grid square * (up to) 5 tiles: bare soil (various types), low vegetation (various types), forest (various types), water, ice; in addition: possible snow cover * assimilation of soil surface temperature and soil moisture * improved turbulence scheme Questions/Comments: ------------------- (N. Gustafsson/H. Schyberg): - Consider using SAF sea-ice product for HIRLAM/ISBA. - Extraction of 2-year data set done by met.no; EUMETSAT AMSU retransmission: set up data flow in about 6 months. (all): - Some discussions on "incest problems" that will probably be of relevance to us. In particular, using ice analyses from the day before eliminates model error correlations. Also a point on bias correction of satellite data, which ECMWF only performs close to ground based stations; HIRLAM will probably follow that approach. 2.3 Part 3: Empirical Model f. Emissivity and Backscatter of Sea Ice (L.Toudal) =============================================================================== - Part 3 done by DTU-DCRS - Importance of ice signatures ("Is the signature due to ice concentration or due to ice surface type?") - QuikSCAT: HH-VV-ratio clearly shows ice edge - backscatter: higher for multiyear (MY) ice (less saline => more penetration => more scatt. by inhomogeneities) - emissivity: higher for first-year (FY) ice - example QuikSCAT time series of backscatter, 2002: * stable in winter * highly variable in summer, caused by variable surface properties: wet snow, surface melting * in July, MY ice not visible because backscatter reduced dramatically (wet snow/ surface melting) - example SSM/I time series of ice emissivity, 1996-97: * stable in winter, but with slow variations, espec.: increase towards end of winter that causes retrieval of erroneously high ice concentrations (>100%); due to surface characteristics like wet snow cover * highly variable in summer * over permanently open water: variable all year, but more so in summer - models: * MWMOD: Surface (sea ice and ocean water) emissivity and atmospheric radiative transfer * backscatter model implemented at DMI * validation with satellite observations (temporal/spatial variability) - theoretical consideration of the retrieval accuracy obtainable depending on the channels included; although assumption of uncorrelated channels may not hold there was a rather remarkable improvement through the inclusion of low frequency channels available on AMSR. Questions/Comments: - ------------------- 2.4 Part 4: Sea Ice Concentration Retrieval (S. Andersen) ========================================================= - Part 4 done by DMI - Ice charting at DMI since 1959 (Greenland area) - 500 SAR scenes per year (Radarsat) - SAR image texture classification - thin ice detection: ratio (SSM/I pol.ratio) / (SeaWinds pol.ratio) - joint experience DMI/met.no (SAF HL[high latitudes] centre), passive microwave: * tie points improved, monthly tie points determined * atmospheric correction: includes wind and TWV * tested a number of ice concentration algorithms (sensitivity to CLW) * investigated sea ice emissivity * multisensor products (ice edge, ice type) * multisensor product based on Bayesian approach (ERS Scat, SSM/I) * operational Atlantic sea ice products (from scat., AVHRR, passive microwave) - Goals within IOMASA: * better accounting for snow/ice properties (from Part 3) * improved use of multiple sensors * use of better atmospheric fields over ice (from Part 1) * improved use of high-resolution sensors (but trade-off with more noise) * validation/calibration Questions/Comments: ------------------- (G. Heygster): Use AMSR instead of SSM/I, or is there too much time lag between AMSR and the other SSM sensors? (S. Andersen): The concept of IOMASA is centered around the requirements of numerical weather prediction; the operational constraints within IOMASA require the use of SSMI (AMSR will be considered but it is unlikely that data will be available with sufficient timeliness within the coming years to be useful in NWP model applications). (L. Toudal): One hour might be o.K.; more than that would be a problem for CLW (rapid changes) but still o.k. for sea ice (slower changes). (G. Heygster): ARTIST algorithm (-> L. Kaleschke) has been preliminarily modified to use AMSR (instead of SSM/I), results seem to be considerably better (resolution). (L. Kaleschke): How to validate ice algorithms that are based on SAR images? (L. Toudal): Difficult. But to validate/test the low-concentration performance, a garantueed ice-free area could be chosen (S. Andersen): High concentration conditions must be considered and the best available and most general information is presently SAR data (low concentration tests have been done several times and algorithm performance under such conditions is well documented). (J. Miao): Maybe use R-factor thresholds determined from algorithm sensitivities to CLW. 2.5 Part 5: Demonstration of Real Time Processing & User Interface (L. Toudal) ============================================================================== - Part 5 done by DTU-DCRS - a user interface was developed in the EU-funded project IWICOS (Integrated Weather, Sea Ice and Ocean Service System), and earlier projects like IMSI (Integrated use of new microwave satellite data for improved sea ice observation, ca. 1996-1999) - IWICOS: Information system for user of sea ice information: * via Internet * satellite imagery: SSM/I, QuikSCAT, AMSU, AVHRR * ocean data: bathymetry, waves * weather data: wind * geography data: coastlines, grids * ice data: charts, concentration contours (isolines) - the IWICOS interface is a Java applet - covers potentially ice-affected areas of N and S hemisphere - static data (coast lines etc.) are installed locally - dynamic data, for users on ships, received via satellite (Inmarsat) - the IWICOS interface can be used in IOMASA for dissemination of data to the public user community and for data exchange between project partners - http://www.dcrs.dtu.dk/sea-ice - but: not suitable for internal raw data exchange - we have to find a way for raw data exchange, maybe simple ftp server? - "data broker" server system that was used for raw data in IWICOS (developed by a Finnish company) is nice, but bulky and difficult to maintain, therefore not advisable for IOMASA Questions/Comments: ------------------- (L. Toudal) We can reuse the IWICOS infrastructure (IWICOS funding ends Dec. 2002). (G. Heygster): Will IWICOS java browser run for the next 3 years? (L. Toudal): Yes. It may even be renamed the IOMASA browser. 3. Review of Phase 1 ===================== Ready to start Phase 1. 4. Action Items: ================ 4.1 Two-year Offline Investigation Period and Area ================================================ - What (which data) from whom: * AMSU-A,B: 2000-2002 at IUP (level 1B, level 1C can be generated easily) * Radiosonde, synopt. data: DMI has access to archive * SSM/I: DMI has archive. Restrict to 1 out of 3 sensors (else too much data)? BUFR format - DMI has reading software. * QuikSCAT: at DMI * SAR: ENVISAT preferable over Radarsat because of multipolarisation (or global monitoring mode instead of multipolarisation?) * NWP: met.no/SMHI can supply 20km resolution from Jan. 2003 (just switching from 40 km to 20 km now) - Where (which area): * Overlap region of met.no and SMHI NWP area - When (which time period): * NWP data: 2003-2004 (since better resolution (20km) NWP starts Nov.2002) * remote sensing data: 2001-2002, then 2003-2004 - Which Grid: HIRLAM of met.no and SMHI uses rotated lat-lon grid (equator runs through forecast area). SMHI provides a conversion routine to (and from) conventional lat-lon grid. 4.2 Algorithms ============== - IUP provides Fortran code for TWV algorithm to SMHI and DMI 5. User Advisory Group (UAG) ============================ - potential members of the UAG to contact [by whom]: (a) weather forecast: * ECMWF: Jean Noël Thépaut [T. Landelius / N. Gustafsson] or Erik Andersson (also in SAF) * UK Met. O. (NWP SAF): Steve English [G. Heygster] * HIRLAM: Heikki Järvinen (FMI) [T. Landelius] (b) ice charts, heat flux: * SMHI climate modelling: Markku Rummikainen [T. Landelius] Ralf Döscher? * DMI ice service: K. Hansen [S. Andersen will ask colleagues] * met.no ice service: Helge Tangen [H. Schyberg] (c) heat flux modeling: * ?? [T. Landelius will look for people] 6. Social Objectives ==================== 6.1 Publications ================ - measure for the success of the project - joint publications or publications by one partner 6.2 Technology Implementation Plan (TIP) (C. Melsheimer) ========================================================== - promote and facilitate exploitation of R&D results in FP5 projects - contractual obligation - standardized set of datasheets to be filled in and submitted as draft TIP after one and two years, final TIP to be submitted at the end of the project. - online (electronic) form - "eTIP" - main sections of TIP are about results of the project; several standard categories of results: * usable outside the project * usable in follow-on projects * only usable within the project - most important: identify possible results of IOMASA. Starting point: list of deliverables in the DOW (Description of work), but also other things - main work done by project coordinator (eTIP account registration, filling in project details etc.) - project partners responsible for details on results related to their specific part - C. Melsheimer will supply the partners with relevant information (selected from two big web sites dealing with the TIP: www.hyperion.ie/TIPWebsites.htm and etip.cordis.lu/) - everyone should think about possible results so that we can come up with a draft list of results before the next meeting (in about 6 months, see below) - TIP will be on the agenda of the next meeting. 7. EU OF (operational [ocean] forecasting) Cluster ========================================== - Meeting of EU OF Cluster and 3rd EuroGOOS conference in Athens, 3-6 Dec, 2002 (http://www.eurogoos2002.gr/): G. Heygster will represent IOMASA there. Contributions needed: * about 1 page of text (based on the DOW) and one graphic from each partner, within two weeks. * there will be proceedings (reviewed), deadline around January - Meeting "European Research", Nov 11-13, Bruxelles (http://europa.eu.int/comm/research/conferences/2002/index_en.html): We could only have presented a brochure, for which it is too early. 8. Project Management ===================== 8.1 Reports =========== 8.2. Meetings ============= - Kick-off Meeting (KOM) - 4 Progress Meetings (PM) - one Midterm Review (MR) - one Final Presentation (FP) Month Type Date Place Details ----------------------------------------------------------------------- 1 KOM 11/2002 Bremen 6 PM1 4/2003 København at DMI, Thu/Fr 10-11 Apr. 2003 (tentatively) 12 PM2 10/2003 Norrköping? 22 MR 8/2004 Oslo? 28 PM3 2/2005 TBD 32 PM4 6/2005 TBD 36 FP 10/2005 TBD 9. Any Other Business ===================== - IOMASA Logo: G. Heygster has drafted a first version, will circulate it later. Ideas welcome - IOMASA web site: public area (already set up: www.iup.physik.uni-bremen.de/iuppage/psa/2001/14iomasa.html) and member area (for project partners) will be set up within the next 6 months. - The Contract (EVK3-CT-2002-00067) has been received by University of Bremen, copies have been handed out to all project partners - University of Bremen administration needs account information from the partners in order to forward the advance received from the EU to the partners. - University of Bremen needs an administrative point of contact for each partner Appendix: List of All Action Items ================================== No. Who What -------------------------------------------------------------------------- 1 all identify possible "results" (in the sense of the Techn. Impl. Plan) 2 all (except IUP) tell IUP account information for forwarding of the project money by University of Bremen(bank account no., bank, reason for payment) 3 all (except IUP) tell IUP their administrative point of contact 4 IUP provide AMSU-A,B data 2001-2002 5 IUP provide Fortran code of TVW algorithm to SMHI and DMI 6 IUP set up member area of IOMASA web site 7 IUP (C. Melsheimer) send all partners relevant information on TIP 8 IUP (G. Heygster) contact Steve English about UAG 9 DMI provide radiosonde and synopt. data 2001-2002 10 DMI provide SSM/I data 2001-2001 (which satellite? Communication needed on that. 11 DMI (S. Andersen) contact colleagues (ice service) about UAG 12 met.no/SMHI collect NWP data 2003-2004 13 met.no (H. Schyberg) contact Helge Tangen about UAG 14 SMHI provide conversion routine from HIRLAM grid to conventional lat-lon grid and back 15 SMHI (T. Landelius) contact Jean Noël Thépaut (or Erik Andersson), Heikki Järvinen, Markku Rummikainen (Ralf Döscher?) about UAG 16 SMHI (T. Landelius) look for people for UAG that are interested in heat flux modeling Acronyms/Abbreviations ================================= AMSU - Advanced Microwave Sounding Unit (part of ATOVS) ATOVS - Advanced TIROS Operational Vertical Sounder CLW - cloud liquid water (Euro)GOOS - (European group to support the) Global Ocean Observing System FY - first-year MY - multiyear NWP - numerical weather prediction OLS - Operational Linescan System, Vis/IR sensor of DMSP SAF - Satellite Application Facility TBD - to be defined TIP - Technology Implementation Plan TWV - total water vapour UAG - User Advisory Group Minutes taken by Christian Melsheimer University of Bremen Institute of Environmental Physics PO Box 330440 28334 Bremen,Germany Tel.: +49-(0)421-218-3041 Fax: +49-(0)421-218-4555 e-mail: melsheimer@uni-bremen.de http://www.sat.uni-bremen.de/ Some additions and corrections by N. Selbach, L. Kaleschke, G. Heygster and S. Andersen. ---------------------------------------------------------------------- LocalWords: IOMASA IUP EU WP CLW TWV AMSR KOM eTIP FP ECMWF HIRLAM EUM SAF EC LocalWords: NWP EuroGOOS Künzi Justus Notholt Georg Heygster Melsheimer Miao LocalWords: Junggang Nathalie Selbach Kaleschke DTU DCRS Leif Toudal DMI SMHI LocalWords: Søren Harald Schyberg Tomas Landelius Nils Gustafsson RS PELICON LocalWords: SSM GHz DNMI OLS AMSU MODIS ATOVS NESDIS BUFR HRPT EUMETSAT ISBA LocalWords: HH VV Radarsat IWICOS København Norrköping EVK CT Tangen DWD FY LocalWords: Kalman RTTOV QuikSCAT multisensor Inmarsat multipolarisation UAG LocalWords: Noël Thépaut Andersson UK Heikki Järvinen FMI Markku Rummikainen LocalWords: Ralf Döscher Helge html TVW GOOS Linescan