HydAlp
Minutes of the 1st Technical Meeting held at the
Institut für Meteorologie und Geophysik,
Innrain 52, A-6020 Innsbruck, Austria,
3rd/4th March 1997
Chairman: Helmut Rott
Agenda
Welcome
Approval of agenda
Overview of the software systems at the institutes (graphic, image processing etc.)- All
Discussion on software standardisation/convergence. - All
Overview of the possible test basins - IMGI, MLURI, SMHI, UBE
Summary on available hydrological models - SMHI, UBE
Planned involvement of customers - VERB,MLURI, IMGI, et al.
Presentation of preliminary remote sensing data base - IMGI, MLURI, SCEOS, UBE
Overview on conventional data base - IMGI, MLURI, SCEOS, UBE
Review of individual tasks for month 1 to month 9 - WP co-ordinators
Action items
Next meeting
AOB
List of Participants (in alphabetical order)
Josef Aschbacher (on 3.3.97) EC-CEO
Michael Baumgartner UBE
Chris Clark SCEOS
Justin Morgan-Davies MLURI
Rob Ferguson SCEOS
Graham Glendinning IMGI
Barbro Johansson SMHI
Hannes Kleindienst UBE
David Miller MLURI
Mats Moberg SMHI
Thomas Nagler IMGI
Otto Pirker VERB
Shaun Quegan SCEOS
Erich Riegler DIBAG
Helmut Rott IMGI
Gary G. Wright MLURI
Acronyms
AVHRR Advanced Very High Resolution Radiometer
BASAT Test basin in the Central Alps of Austria
BASCH Test basin in the Swiss Alps
BASSW Test basin in Northern Sweden
BASUK Test basin in Scotland
ERS European Remote Sensing Satellite
DBMS Data Base Management System
DEM Digital Elevation Model
HBV Hydrological Model of SMHI
HROI High Resolution Optical Imaging Sensor
MROI Medium Resolution Optical Imaging Sensor
NOAA US National Oceanic and Atmospheric Administration Satellite
RESP Responsible
SAR Synthetic Aperture Radar
SIR-C/X-SAR Spaceborne Imaging Radar-C/X-SAR
SRM Snowmelt Runoff Model
1 Welcome
At 16:30, Dr. Rott welcomed the partners to the 1st Technical Meeting of HYDALP. Copies of the previous sessions overheads and blanks for the division of responsibilities in the various Work Packages (WPs) were given to each participant.
2 Approval of agenda
The agenda of the 1st Technical Meeting was approved by the participants.
3 Overview of the software systems at the institutes (graphic, image processing etc.)
It was agreed by all participants that the use of GIS and DBMS packages by each group would continue, with the proviso of a complete description of the input and output protocols and inter-package conversions. The format of parameter storage within databases must similarly be documented.
The problem of remotely running a GIS was mentioned by Dr. Baumgartner, and it was thus decided not to set up a central GIS. He also discussed the use of SRM running with DBASE, the PC database.
The two snow models HBV and SRM are PC based, and it was decided not to attempt a system move to Unix. Once more, it was highlighted that a comprehensive set of data transfer protocols would minimise problems associated with different platforms.
In this context, programming languages were discussed. No languages were given preference, but it was highlighted that the code should be as modular as possible.
4 Discussion on software standardisation/ convergence.
It was decided that the CEO needed to define the interface for the final demonstration package. SRM and HBV would remain stand alone models, with strap on modules available. These modules would be in a standard form, as would their documentation.
It was also concluded that the models needed to be fully described and documented at an early stage, along with parameter lists, units and ranges. This was illuminated by a presentation from Dr. Miller of MLURI on organising information on data sets by compiling tables of meta data.
5 Overview of the possible test basins
BASAT
Dr. Rott (IMGI) firstly discussed the Austrian Basins, presenting the work of the IMGI in the Öztal. The DEM, and field stations were described, and the large errors in estimating the areal precipitation highlighted. This was followed by Dr. Nagler’s (IMGI) discussion of the Zillertal/Gerlos, and previous analysis using SAR and SRM for runoff simulation. Rob Ferguson pointed out that the variety of the hydrological regimes in the project was a great strength. With no glaciers in the Scottish basin and only small percentages in the Swedish basin, this could be improved with a basin in Austria with a larger proportion of glaciers.
Information on the Austrian basins are given in Appendix 1.
No decision was made about the final choice of the Austrian basin.
BASUK
Dr. Wright (MLURI) then gave an overview of the Spey catchment in Scotland, emphasising the diversity of water user and the importance of conservation in the area. It was also mentioned that large quantities of water would be taken out of the system by a predicted reforestation of 400%. Attention was then drawn to the points likely to be of interest to the Scottish Customer Focus Group.
Dr. Ferguson (SCEOS) discussed the geography and complex hydrology of the Spey catchment, giving a history of hydrological research and interest in the region. Real time meteorological data is available in the ski areas. He also outlined previous snow hydrology work in the Cairngorms.
The special features of BASUK are summarised in Appendix II.
It was agreed that this was the only choice for the Scottish basin, but that a sub-basin was yet to be decided, so as to emphasise the effects of snowmelt.
BASSW
Dr. Johansson (SMHI) discussed the choice of Swedish basin Litnok. She emphasised the inaccessibility and remoteness of the site near the northern border with Norway. There are 3 synoptic weather stations near to the basin, but no meteorological data collected within it. Meteorological forecasts up to 48 hours are made with Hirlam (high resolution limited area model), which is actually a co-operation between Sweden, Norway, Finland, (Denmark?), Ireland, Spain and maybe some other country.
A DEM was available, but permission was to be sought for SCEOS also to use it.
The Swedish basins are given in Appendix III.
A final choice on the basin and sub-basins was not made, although Litnok is a good candidate.
BASCH
Dr. Baumgartner (UBE) then went on to discuss the Swiss basins, and the Alpine Snow Cover Analysis System in operation. He drew attention to the use of online AVHRR data, the DEM, the use of up to 10 aspect classes, and a single meteorological station for the runoff model. This is all undertaken within a GIS. Mention was made of the WMO intercomparisons of runoff models (including HBV and SRM) undertaken in the Swiss Alps.
The Swiss basins are given in Appendix IV, and a flow chart of the Alpine Snow Cover Analysis System in Appendix V. No decision was made regarding the choice of Swiss basin.
6 Summary on available hydrological models -
SMHI, UBE
HBV model
The hydrological snowmelt runoff model, HBV was described. It originated in the 1970s, based on the degree-day method, and requires inputs of temperature, precipitation and evapo-transpiration. It was utilised to forecast runoff for hydropower, and recently re-evaluated with satellite imagery. It is PC based, and functions with up to 300 sub-catchments, and each sub-catchment can be devided into a maximum of 50 elevation-zones. A multi-year record is required for calibration (e.g. 10 years). The model takes into account to forest cover. The model is described in the user manual. HBV documentation was provided for the participants.
SRM model
The Snowmelt Runoff Model was then described by Dr. Baumgartner (UBE). It is also a degree-day model, and run according to depletion curves determined from previous imagery. The inputs are temperature, precipitation, and snow cover from satellite imagery. There is no input for evapo- transpiration. It is semi-distributed, and may have up to 8 elevation zones. The degree day factor is calculated on a biweekly basis, and the runoff coefficient given by tuning previous model predictions in a given basin. It was pointed out by Dr. Rott that the runoff coefficient and the degree day factor are the crucial factor for accurate forecasts, and are not known on a new basin. The model is described in the user manual. SRM documentation was provided for the participants.
A question was raised by Graham Glendinning concerning modifications to the snow models, and whether a given version was to be chosen. Dr. Rott emphasised that all modifications should be made in a modular manner, so that both different basins could be compared, and new techniques investigated.
7 Planned involvement of customers -
VERB,MLURI, IMGI, et al.
DI. Otto Pirker, the representative of VERB, outlined the scope of the work of the VERB, and highlighted the need for short-term runoff forecasts, for sufficient time to staff and power up the generator stations. Any improvement to this was seen as important to the VERB due to the large quantities of electricity derived from stream generation. He also drew attention to the wide spatial range of concern to the VERB.
8 Presentation of preliminary remote sensing data base - IMGI, MLURI, SCEOS, UBE
The availability and cost per scene of remote sensing imagery was discussed. Medium resolution sensors mentioned were AVHRR and Resurse. High-resolution sensors included Spot, IRS-C, Landsat TM, Early Bird and MOMS. The availability of SAR and Radarsat imagery was also highlighted.
It was decided that the remote sensing data search of SCEOS would include BASUK and BASSW, but not BASAT or BASCH. This was due to the comprehensive data sets already available at the Austrian and Swiss institutes.
9 Overview on conventional data base - IMGI, MLURI, SMHI, UBE
This item was covered in section 5, by the respective partners.
10 Review of individual tasks for month 1-9 WP co- ordinators
The following duties were mentioned. All of the WP duties were not covered, but many were addressed in the kick-off meeting, and all are outlined in the Technical Annex.
WP210: Swiss customers should also be contacted
WP220: Every basin was to have its own database, as the work involved in moving everyone onto the same platform was not feasible within the given time frame. This necessitates rigorous documentation for the data format information and transfer.
WP230: Remote sensing data search: SCEOS, UBE, and IMGI. The choice of basins and melt seasons was to be completed early, so as to enable the remote sensing data search (see Action Item 3.). Sensors and costs were also to be investigated.
WP240: Field experiments were discussed, and the possibility of joint fieldwork was to be deliberated at the next Technical Meeting.
WP620: It was decided that the WWW pages introducing the project would be held at the CEO, with links to the various partners. Private WWW pages for internal project use are located on WWW sites of the partners. For the exchange of the documents the HYDALP ftp-account at CEO can be used.
At this stage, David Miller gave a demonstration of a WWW site for a different project conducted by MLURI and others, highlighting the possibilities for the HYDALP WWW site.
11 Action items
Item No.
Action
Responsible
Deadline
1
Collect information on computing platform, languages, data formats, image processing, graphics, GIS, and data base management. If applicable, a documentation standard should be suggested.
All
20th March 1997
2
Hydrological model distribution: SRM at http://hydrolab.arsusde.gov/gi-bin HBV: Software and Agreement forms
SMHI
30th April 1997
3
Define basins and melt seasons
IMGI, MLURI, UBE, SMHI
30th April 1997
4
Select DBMS
IMGI, MLURI, UBE, SMHI
15th April 1997
5
Install Project Summary on the WWW at the CEO
MLURI
30th June 1997
6
Design a project logo. Winning design wins a bottle of Schnaps, courtesy of Helmut Rott. Needs to be available for Action Item 5.
all
30th June 1997
12 Next meeting
The next meeting was set for 17th/18th November 1997, in Bern, chaired by Dr. Baumgartner (UBE).
13 AOB
At 15:00 the 1st Technical Meeting of the HYDALP was declared closed by the chairman, with SCEOS and SMHI then leaving.
The remaining participants adjourned to an adjoining room, where Otto Pirker discussed further the needs of the VERB, with explanatory diagrams showing the production of power over the year in Austria, and the comparatively high proportion of hydropower. Dr. Miller asked about a cost benefit type of analysis, and comparisons with previous methods. It was stated that the VERB required the model predictions to be better than the brain of the customer, as present methods relied on human interpretation and experience.
Referrences:
Sten Bergström, 1992, The HBV Model- its structure and applications, SMHI REPORTS HYDROLOGY, RH No. 4, April 1992
Martinec J., Rango A., Roberts R., 1994, Snowmelt Runoff Model (SRM) User’s Manual, ed. Michael F. Baumgartner, updated edition 1994, Version 3.2, Geographica Bernensia P29.
Minutes of the 1st Techn. Meeting- HydAlp.