Present
Ole B. Andersen (NSI), P. Bauer-Gottwein (ER-DTU), Rene Forsberg (NSI), P. Knudsen (NSI), G. Strykowski (NSI), Mike Butts (DHI), D. Rosberg (ER-DTU)
Agenda
1) Welcome
2) Hydrograv status – budget and project
3) Short introduction of partners (please limit to 5 min)
4) WP1 - GRACE Data - Ole
5) WP2 - Forward Gravity Modeling – Peter
6) WP3 - Gravity Surveys – Peter, Ole, All.
7) AOB, Next meeting, Project web pages
Summary of relevant points:
Buying an Scintrex CG5 instrument
Olivier Francis has a CG5 in Luxemburg and will be consulted. This instrument was employed in Godhavn for a month this summer for the study of loading tides. The instrument can work autonomously for weeks/months. We also have a feedback system to an old Lacoste Romberg gravimeter. This should be able to give nearly the same accuracy as the CG5 over time.
Pumping test (groundwater from gravity)
One suggestion is to make a groundwater pumping test at a well. MB has the details of where these are performed. Pumping is expensive and the gravity observations would give information on the movement of water in connection with such pumping. The spatial and temporal signal associated with such a pumping will indicate the porosity and structure of the sediments.
The pumping would also be interesting to demonstrate that time-lapse gravity works and can give signals and as such make the demonstration case in Denmark as mentioned in our application. The whole pumping exercise could also be a good Masters study and could be published in hydrological journals.
The setup should use 2 or 3 instruments away from the pumping. A pumping is expensive and gravity gives cheap info on the permability of the soil etc. The suggested time is late 2007 for a pumping test. Location to be determined by MB in connection with Mads Møller at COWI.
Instrument calibration
It was suggested to calibrate the instrument using the SG observations of the permanent GPS sites in Jutland in Smidstup and Suldrup. Both should have SG observations, and we could investigate the properties of the instrument by re-measureing the difference at the two sites two to three times.
Places to be used for calibration:
Vestvolden is the new Danish Buddinge point and has lots of repeated SG observations and there is also a well next to the reference point. GS should be involved in the observations and sharing of data with .
Wettzell (Germany). We could consider making a test of the instrument in Wettzell, where we have accurate measurements. Strasbourg (France) – Ole to contact Jacques Hinderer on data from the SG observations and observations in Africa.
Okawanga survey
One suggestion is to carry out a survey of maybe 5-10 stations and then let the instrument stay on the central station provide the time-lapse gravity signal. Subsequently, a second spatial survey should be carried out and the temporal signal should be estimated again and perhaps one final spatial survey.
It is important that the spatial surveys are carried out at the exact same repeated points. One cm of height differences gives 0.3 uGal, and we also have to account for pressure.
The weather station can be provided by the water authorities in Botswana. The problem of GRACE measuring a much larger area can be addressed by trying to model the changes in the surrounding regions using hydrological models from i.e. GLDAS.
Peter and Silvia had found out that the Okawanga delta signal itself on GRACE, due to the horizontal water movements, is only of the order of 0.5 uGal if only the Saturated Zone is considered, and thus smaller than measurable by GRACE alone. An estimate of the gravity effect of soil moisture in the Unsaturated zone surrounding the delta will be studied.
Employing two Ph.d. candidates
It was suggested to put out an ad in EOS on the two Ph.D offered in the project. This is requested by forskningsråded (Per thought) and we need highly qualified candidates.
Software for estimating gravity from water storage
Rene has software for the calculation of prism effects which use a standard technique in geodesy and is directly applicable in hydrology. Peter got a copy of the GRAVSOFT manual.
The in-finite disk is problematic to use and should be avoided. It is just used on regular base because of the ease of use, but is far from accurate.
Peter estimated that 90% of the signal from a finite disk lies within 10 times r – the height of the observations. For GRACE this means that within a disk of 4500 km.
GRACE
The timelapse gravity from GRACE is a seriously complicated task and requires a lot of effort and international corporation. Perhaps a combination technique could be used where the long wavelength are estimated initially and removed and the finer scale signature is resolved from the residuals.
There are also routines in GRAVSOFT that can be used for calculating forward response of mascon approaches. This might be a possibility. Gabriel would be very interested in trying his forward modeling technique on the Okawanga delta data.