DRAFT Minutes of the 2nd Technical Meeting held at University of Bern, 17-18 November 1997Author: Thomas Nagler, Graham Glendinning AGENDAChair: Helmut RottDay 1: Monday, 17 Nov 97, 9:00 h s.t.
List of Participants:
1. & 2. Welcome and approval of agendaThe meeting was opened at 9:00 am. M. Baumgartner welcomed those present to the Geography Department at Berne. The Agenda was distributed. Point 8 - WWW Demonstration of the Agenda was shifted to day 2, so that it could be presented to the representative of EC-CEO J. Aschbacher. During the meeting several times the Technical Annex was addressed. For clarification the Technical Annex to the contract for the HYDALP Project, dated 12 November 1996, is the binding document for EC and it supersedes earlier versions. 3. & 4. Summary of work performed - Report on ongoing tasks. 200-Baumgartner, 300-Quegan, 400-FergusonAll participants agreed to integrate agenda items 3 & 4 to save time. 200 Assessment of Customer Needs and Database CompilationAn overview of the work in WP200 was presented by the WP co-ordinator, M. Baumgartner. The interim report, RM-2, was ready for review, and databases were documented. It was emphasised that databases are always under development as data will also be acquired for the melt seasons 1998 and 1999. 210 - Specification of Customer Needs, M. BaumgartnerThe report RM-1 (Interim Report on Customer Requirements) was discussed. H. Rott stated that the interim reports are to be considered flagships of the project. The draft report had been reviewed and required some additions to set it in a wider European context. To accomplish this aim, contributions from partners in other countries regarding additional users were requested. The report RI213 by O. Pirker was also marked for inclusion. B. Johansson mentioned a hydropower company in Sweden, which might provide information. M. Baumgartner mentioned that the Swiss Hydropower companies may contribute to the customer reports. He outlined the problems to get in contact with the Swiss hydropower companies. He also covered problems in modelling the runoff within the Swiss basin due to the storage and release of water and pumping water between reservoirs. The reservoirs were often co-owned by several concerns, increasing problems of data access. Online information of runoff data are not available. This information was vital for improved runoff simulations, and especially for future forecasting. It was hoped that contact with the largest of the companies would bear fruit in the following month. 220 - Hydrological and Meteorological Data Base CompilationH. Kleindienst presented the objectives of this work package as outlined in the Technical Annex. The data management of the partners was described, and is given in detail in RM-2. From the discussion it was concluded that the cost assessment of meteorological and hydrological data is required. H. Kleindienst identified a number of areas for further work. He expects contribution from each partner:
Regarding WP225 (Data Access Module) it was stated that a report RI110/225 was completed. The data access module was covered in detail in Agenda Item 12. 240 - Field ExperimentsT. Nagler presented the proposals for fieldwork in BASUK and BASAT in 1998, which will be organised by IMGI and MLURI respectively. SCEOS and UBE may participate in one of the field campaigns. Field measurements will concentrate on a limited set of parameters, which are needed for verification of remote sensing products, and additional meteorological / hydrological parameters. The locations and the timing of field measurements and the parameters, which will be measured, were presented and discussed. In 1999 field campaigns are not planned. 300 Remote Sensing Methods and AnalysisS. Quegan presented a summary of WP300, its progress to date and the problems it has encountered. The objectives were given, as outlined in the Technical Annex. The overall status was described. It was emphasised that it is the remote sensing component that interests CEO, and indeed that the future of CEO also depends on the satisfactory results of such projects as HYDALP. S. Quegan gave an overview about the WP 310, 320, 330. The report RI311, review of methods and improvements, is in delay, requiring the addition of a more detailed remote sensing component. This task impacts upon the internal tasks WP312 - WP315, external report RM-3, and work packages WP340 and WP420. No further problems were anticipated in the completion of this task. Despite this, the tasks WP312 - WP315 were on schedule. WP320 (remote sensing methods for high resolution sensors) is in time. Work was done mainly for developing the RS.SAR module. The work for processing HROI data needs further specifications. A discussion paper on geo-coding has been completed and will form the basis of RI322 (delivered in month 10). It was decided that RI322 will cover only geocoding of SAR and HROI data. The topic "information extraction from remote sensing data" will be included in RI321. SCEOS will supply this topic for inclusion in RI321. WP331 (Remote Sensing Methods for MROI) is severely delayed, which also impacts upon WP302, WP340 and several activities within WP400. The initial results of WP332 are unpromising. WP333 started this month, and no progress has been made as yet. It was decided to transfer expertise of one partner to other partners as much as possible. Within HYDALP all software will be shared between partners. Property rights of software and data for use after HYDALP will be an agenda point in the next Technical Meeting. The discussion turned to the topics to be included in RM-3. RM-3 will provide a coherent overview of the work on remote sensing methods conducted in WPs 310, 320, and 330. The problem with the tight schedule was emphasised, and it was made clear that the timely delivery of reports RI321 and RI333 is crucial. It was decided that a draft of RM-3 would be delivered on 1 March 1998. At this time information from WP331 has to be provided by UBE to be included in this report. Further information may be added as an Appendix. WP340 was then outlined, as described in the Technical Annex. 310 Extraction of basin characteristics, D. MillerThe objectives of the work package were presented as given in the Technical Annex. Basin characteristics were split into core and non-core divisions, and the availability of each in each basin was given. Mention was made of the importance of the inclusion of tolerance, variability and error information throughout the data flow process to the user. The sensitivity of hydrological models to land cover information was discussed. Required information should be provided by hydrological modellers. BASUK and BASAT were chosen as basins for land cover classifications by means of remote sensing within the demonstration package. 320 Remote sensing methods for high resolution sensors, T. NaglerFor the tasks WP321 and WP322 the partners IMGI and SCEOS are working in close co-operation. The objectives of the work package were presented as given in the Technical Annex. Work was focused on the development of the RS.SAR module, from which the design and status was presented. An overview of the preliminary processing of HROI data was given. Available software packages for extraction of parameters from HROI data were discussed. Geo-coding methods were discussed at some length. These will be included in RI322. Foreshortening information will also be included. 330 Remote sensing methods for medium resolution sensors, S. VoigtSince 15 October 97 S. Voigt is employed at UBE. He is responsible for WP330. The delay of WP331 and WP333 were discussed at length, as they may present a serious threat to project objectives. The reason of the delay was a shortage of manpower at Bern. After consultations with the CEO representative (J. Aschbacher) on the 2nd day, it was decided not to delay the production of WP333, as this information is required for the external report RM-3. The MROI module was pushed back to 1 April 1998. The objectives of the work package were presented as given in the Technical Annex. The literature review for WP331 is complete, but further progress not yet been made, delaying other work packages (WP441). S. Voigt gave an overview about the present method for snow mapping using NOAA-AVHRR data at UBE. The procedure was said to take one full day of interactive analysis per scene. The results of the following discussion was, that a semi-automatic method for medium resolution image analysis has to be developed, including a set of objective classification rules. This module can then be provided to the partners. It was decided to develop the module in ERDAS; if required, it could then be converted for PCI. A commonality of training sets was required, perhaps with the use of a reference image. The priority for development of MROI modules (WP330) was set as follows:
B. Johansson mentioned the need of maps of snow coverage for BASSW from previous years for HBV model development (WP 441). UBE agreed to undertake a May-July search for historic (in the last 10 years) NOAA AVHRR imagery covering BASSW (which is covered by the same AVHRR path as BASCH) and supply the snow maps to SMHI. Historic AVHRR data are archived at UBE. A high resolution image will be analysed by O. Turpin for land cover classification. The results of WP332 (M. Broatgate), water balance and evapo-transpiration, were discussed. There was no obvious relation found between NDVI based on AVHRR data and evapo-transpiration in BASUK. The conclusion from the discussion was that the data set is not appropriate for these investigations. Thus, report RI332 should include only a literature review. 400 Hydrological Modelling, R. FergusonThe modellers meeting, which took place in Bern on 17 November 1997, was reviewed and a summary of the draft report, RI411 on SRM-HBV inter-comparison, was given. Completed and future tasks were described. Ideas on possible model modification were presented by R. Ferguson, as described in the minutes of the Modellers meeting of 16th November 1997. 420 Definition / calibration of model parametersThe present state of tasks in this work package was outlined by G. Glendinning. The HBV model has been installed at SCEOS, IMGI and SMHI, and the SRM model at IMGI and UBE. B. Johansson promised a document to aid calibration. Delays in WP422, 423 and 424 were not considered a problem for other work, and these tasks are to be completed by 15 January 98. Further calibration was also needed as part of WP421, before the parameters are compiled. 5. Remote Sensing data base / ordering data sets for various basins (Caves, Nagler, Baumgartner)O. Turpin presented proposals for image acquisition in BASUK and BASSW. Funding was to cover past data for SRM depletion curves, and future data requirements. It was noted that SRM would not be applied to BASSW, so the historical imagery is not required. BASUK will require the generation of depletion curves, and three such curves are planned (for heavy, average and light snow years). Six scenes were recommended covering the snow depletion. The data acquired to date were listed, further data needs were given. Only two high-resolution optical images (one per basin) could be bought per melt season due to financial restraints. MLURI would acquire one or two further images if possible. One MSS image will also be obtained per basin per melt season. MROI would be available as AVHRR images from Dundee or Bern free of charge, ERS-SAR imagery will also be acquired. The detailed costs for image acquisition were presented. H. Rott emphasised the importance of testing the capabilities of SAR wet snow mapping in BASSW and BASUK, before a lot of SAR scenes will be ordered. The preliminary investigation of the capabilities of SAR wet snow mapping in BASSW and BASUK has to be completed as soon as possible, to learn about the applicability of this data. T. Nagler presented the data acquisition plans for BASAT in 1998 and 1999. M. Baumgartner discussed the ongoing acquisition of the AVHRR receiving station, and the future purchasing of a RESURS receiving station. This is planned to be operational in the early summer of 1998. RESURS MSU-SK and NOAA AVHRR data will be provided free of charge within the HYDALP project. The uncounted costs of storing and viewing imagery were highlighted by B. Johansson. 6. ERS 3rd Announcement of OpportunityH. Rott announced the possibility of obtaining ERS SAR imagery from ESA through a proposal in response to an announcement of opportunity. Proposals are due on 31 January 1998. He encouraged others to apply for imagery. IMGI is unable to respond to this announcement as the ERS image acquisition in BASAT is covered by an ongoing ESA AO-experiment. S. Quegan agreed to take the lead role on this for HYDALP within the basins BASUK and BASSW. 7. Selection of basins and data sets for intercomparisons of MROI - HROIH. Rott opened a discussion on the choice of basins for comparison of high and medium resolution methods. It was decided that one alpine and one non-alpine basin would be chosen. After discussing pro's and con's of each basin, BASSW and BASAT (extended to Zillertal/Pegel Hart, basin size 1050 km2) were chosen, implying an addition to the workload at IMGI beyond the work allocated in the Technical Annex. M. Baumgartner agreed that the medium resolution imagery (NOAA AVHRR, Resurs MSU-SK if available) of BASAT would be supplied and processed by UBE, as Zillertal is covered by the same swath as BASCH. The set of NOAA-AVHRR imagery for BASSW would be completed by SCEOS (Dundee receiving station), as planned. The timing of the inter-comparison of MROI-HROI was given as the middle of 1998 (draft report in June 1998), under WP340, with contributions from WP320 and WP330. 8. WWW Activities (WP 600: Miller)M. Broadgate presented the planned work regarding the WWW activities. In line with the decisions made at the 1st Technical Meeting, simulated model runs will be pre-prepared and available on the WWW as pre-defined options. This enables the models to be run in advance and avoids giving open access to any copyright software. D. Miller gave a presentation of the STORMS project WWW pages, as a possible template for future HYDALP work. 9. Presentation of the Alpine Snow Cover Analysis System and visit to the satellite receiving station at UBEA presentation of the Alpine Snow Cover Analysis System - ASCAS was given by M. Baumgartner and S. Heilmann. Then the first day of the meeting was then declared closed 10. Summary of technical achievementsH. Rott summarised the objectives and the success criteria of the HYDALP Project (copies of the viewgraphs are attached). J. Aschbacher pointed out that different views can be applied to determine the successfulness of the project. One view could be, that the project is successful if all deliverables were given at the right time and that all reports are available, but CEO aims beyond that, proving the way for operational remote sensing. D. Miller proposed drawing up some points for assessment of success criteria, e.g. model quality and cost effectiveness. O. Pirker mentioned that for customers the benefit of using remote sensing data is crucial. For him remote sensing provides new information and is supplemental to other measurements. J. Aschbacher suggested to include success criteria in the report RM-1. But H. Rott mentioned that the RM-1 is in draft and it may be too late for this. Regarding the assessment of success of a project J. Aschbacher outlined that CEO has no experience with other projects, because HYDALP is in the first wave of projects. Furthermore he stated that there are non guidelines at CEO. CEO hopes, that the results of a project will be used by a company or that a new company will be founded based on the work within the project. J. Aschbacher also mentioned that a customer within the Water Resources board at CEO might be interested in the results of HYDALP. He will provide information to H. Rott. H. Rott presented the HYDALP achievements from Moth 1 to 8. 11. Management issues, critical points, risk assessmentManagementH. Rott stated that the selected management structure is in principle working well. The co-ordination and communication within the tree structure works fine, but the interaction on technical and scientific issues across branches must be improved. Further improvements are needed for:
Delays of Monthly Task reports occurred mainly in the first 3 months, the last months they were mostly in time. As H. Rott outlined monthly task reports should focus on main points and provide realistic assessment of status. J. Aschbacher mentioned a red-flag procedure at EC, which is started if a project goes into the wrong direction. The red-flag procedure corresponds to an evaluation of the project by independent experts, who conduct a technical and financial audit. M. Broatgate proposed that each task manager gets also information back from higher levels to get a better overview about the work within the whole WP. H. Rott agreed and he stated that a monthly WP Status report (for each WP X00) will be compiled by IMGI based on the Monthly WP reports. It will be distributed to all persons involved in the project. The monthly WP Status report summarises the work done within this WP (highest level e.g. WP 300). Important sources of information on ongoing and planned activities are also the 3- monthly Progress Reports, which are available to every HYDALP collaborator. Risk managementH. Rott presented possible risks within HYDALP. One of the major risks is the limited applicability of remote sensing tools under regional conditions. This is especially the case for SAR snow mapping, which has not yet been proven for BASUK and BASSW. Optical images are less critical in respect of snow cover detection, but the timing may be critical. The availability of time sequences of satellite data should also be checked (HROI for BASUK, BASSW). If suitable time series are not available, alternative schemes for the satellite component have to be worked out. Within the hydrological part of the project it is necessary to focus on the main objectives and to keep the time schedule for development and testing as tight as possible. Otherwise the pre-operational tests are at risk. The preparation of the pre-operational runoff forecasting activities including planning for near real time satellite and hydromet data transmission should start early. Regarding the near real time transmission of SAR Scenes ESA/ESRIN will be contacted by H. Rott. The pre-operational run in spring 1999 is very important for the success of the project. Because of the start date of the project in March 1997 the project will be finished in August 1999. Therefore there will be only little time to analyse and compile the results. H. Rott proposed the extension of the project in terms of time (2 months) without extra money from EC, because the main snowmelt runoff period in BASSW is from May to July and in BASAT from April to June. In order to include the results from the pre-operational tests 1999 in these basins in the final report, the extent of the project is needed. J. Aschbacher stated that this is in general not the problem, if clear reasons are given. H. Rott stated that the present structure of reports (3 monthly progress reports) will be kept. For extending the project in time the project co-ordinator H. Rott has to contact EC DGXII. J. Aschbacher suggested to contact DG-XII as soon as possible. H. Rott outlined that he plans this in early 1998 when some work on hydrological modelling has been done. 12. General Flowline / Data Access ModuleG. Glendinning gave an overview about the General Flowline/ Data Access Module and the tools which were proposed for handling remote sensing products and hydromet data and for reformatting data. One point of discussion was the concept of hydrological units and of meteorological weighting factors. After a while H. Rott stated that the discussion was going nowhere and a decision had to be made. Thus a screening of positive / negative aspects of the concept was made by every participant. A summary of the arguments is presented below:
During the discussion several statements were made, but no real consensus was achieved. It was stated that no other option had been proposed. The modellers were unsure which tools would be most useful for the work in HYDALP. B. Johansson outlined that a digital snow map may be sufficient for SMHI. At the end it was decided, that the development of the tools with full version (as they are described in RI110/225) by DIBAG is cancelled. A trimmed down version of the Flowline would be proposed by IMGI within the next weeks based on the points discussed. 13. Activities of upcoming tasks: 430- Clark, 440-Johansson, 510-Kleindienst, 520-Rott430 - Data Fusion: Remote Sensing and in situ Data (Responsible C Clark)R. Caves presented the work planned within this WP for C. Clark, who did not participate in the meeting. At first possible relevant measurements of relevance were summarised, which include ground based measurements, remote sensing data, and meteorological data and forecasts. The following topics were covered:
Methods and examples for each of these topics were given. H. Rott pointed out that the WP has to focus on methods which are of priority for HYDALP, but a clear plan was still missing. 440 - Hydrological Model Modification (Responsible: B Johansson)B. Johansson presented possibilities how to improve the HBV model. This included the extrapolation of meteorological data (e.g. lapse rate variation, which is probably not important as stated by H. Rott), estimation of the melt rate (based on energy balance), subdivision of basins, and the use of snow coverage based on RS data. Furthermore the HBV model will be modified for the use of distributed data. The snow extent derived from remotely sensed data will be used to update the snow coverage calculated by the model. This includes the development of a program by SMHI. Models, which link snow coverage and snow water equivalent, can be applied for interpolation/extrapolation of remotely sensed snow extent. For alpine basins a routine for glacier melt was proposed. Maps of surface albedo derived from high resolution optical images would be useful for estimation of the degree day factor. Furthermore first ideas on testing and sensitivity analysis were proposed. For the SRM model the optimum frequency of images has to be investigated. H. Rott pointed out that the next and urgent step is to come up with a plan, which parts of HBV will be modified for remote sensing input and a list of properties for model changes. 510 - Runoff Simulation / Water Balance (Responsible H Kleindienst)H. Kleindienst proposed a log-file structure, which reports the runoff simulations of HBV and SRM model. The log-files can be distributed to the modellers or stored on the HYDALP Homepage at MLURI and provides a structure for exchanging experience for runoff simulation. The simulation log-file includes the initial parameters of the model and the description and result of the simulation. It was decided that the log-file will be stored in Word-format. H Rott stated that an exchange of experience with WP 420 and 510 is useful. Furthermore the impact of EO data on runoff simulations has to be investigated. 520 - Runoff Forecasts (Responsible H Rott)For preparation of the pre-operational runoff forecast period in spring 1999 H. Rott stated that the collection of meteorological forecast data is required (for the basins where forecasts are planned). Other important issues, to be clarified well in advance, are real time acquisition of satellite data and hydromet data. 14. Deliverables for project months 10 to 18H. Rott summarised the reports to be delivered in the period month10 to 18. For some of the reports the deadline was changed, because of unexpected problems within the WPs. An overview on the deliverables is given in the appendix. The consortium agreed in the following deadlines for delivering reports (only changes in respect to the Technical Annex are given):
15. Property rightsH. Rott outlined the exploitation of project results. According to the CEO policy the scope of EC projects is to strengthen the competitiveness of the European Industry without interfering with its business interests. J. Aschbacher agreed and added, that EC is not owner of the products of the project (e.g. software developed within the project). The discussion moved on to the use of software developed within the project. It was agreed that during the project a free exchange of software will be practised. In the next technical meeting the property rights of the software after the HYDALP Project will be discussed. 16. Publications / Conference presentationsH. Rott outlined the rules for publications/presentations within HYDALP. All partners agreed to this rules: For technical papers:
For overview papers:
Agreement with EC-CEO
J. Aschbacher stated that it has to be mentioned in the acknowledgements that the work is funded by CEO. H. Rott showed a preliminary plan for the conferences IMGI will attend. Other participants agreed to send information on planned conference activities for 1998 by 1 December 1997, as deadlines for abstracts are close. 17. Communication / ResponsibilitiesH. Rott mentioned that in some cases the communication between partners and to the project management has to be improved. If there are any changes in responsibilities of task / WP managers then the project co-ordinator and T. Nagler / G. Glendinning have to be informed by email. H. Rott specified key contact persons, who will be contacted when critical decisions are needed (e.g. financial aspects). The key contact persons are:
The key contact persons are asked to declare their time of absence (if longer than 1 week) in the WWW calendar at the MLURI HYDALP Homepage. After a short discussion all partners agreed. 18. Next meetingThe next meeting will be on 11/12 June 1998 at SCEOS, Sheffield (confirmed by S. Quegan on 26 Nov 97). 19. AOBNo other businesses were discussed. Summary of Action Items
Revised Deadlines of Deliverables for Month 9 to 18
|
Minutes 2nd Tech. Meeting - DRAFT 2 02 June 1998