IGCP 565 Workshop 2b: From Satellite Gravity Observations to Products San Francisco, December 12-13, 2009

Notes from the Workshop

• Chris Hartnady, Umvoto, South Africa,
• Richard Wonnacott, South Africa
• Chris Rizos, Australia
• Peter Fitch, Australia
• Hans-Peter Plag, University of Nevada, Reno, USA
• Norman Miller, USA
• Richard Gross, USA
• Eric Calais, USA
• Julia Pfeffer, France

Hans-Peter Plag opened the Workshop and welcomed the participants, and each participant gave a brief introduction. Hans-Peter Plag gave an overview of the project (pdf), summarized the achievements, and discussed the challenges.

The societal motivation for the project stems from the global scale of water resource problems, which are rapidly mounting with dire projections for the future. The science and technology motivation has several layers: (1) geodesy observes and models the Earth system in its three pillars, i.e., the time variable Earth's shape, gravity field and rotation, and observations in all three pillars are strongly impacted by mass transport in the global water cycle; (2) geodetic observations of time-variable changes in the Earth's gravity field, rotation and surface displacements (where the Earth's surface includes the surface of the solid Earth, oceans, terrestrial water bodies, and glaciers and ice sheets) have been used to infer changes in water mass distribution; (3) the global geodetic infrastructure includes many ground-based and space-borne elements operated by many different agencies and institutions. This infrastructure is well posed to provide a significant contribution to Earth observation, and, in particular, for the monitoring of the global water cycle.

Given these pre-conditions, the IGCP 565 project has the goal to utilize the geodetic infrastructure for the monitoring of the global water cycle. This infrastructure has demonstrated the potential to detect at an early stage significant changes in water reservoir storage, and it is this potential that the project wants to explore in order to inform society about changes in terrestrial water storage and to support regional water management. The project has the goal to create a forum where experts from relevant disciplines can meet to explore the potential, identify challenges, and develop solutions. The project's approach is to initiate new research projects and/or link to existing projects; to conduct a series of annual workshops, and to interact with relevant programs.

Hans-Peter Plag then reviewed the achievements from 2008 to 2009. For 2008 and 2009, he listed a number of presentations given at meetings organized by relevant programs. The 1st annual workshop held on December 11-12, 2008 in San Francisco brought together 15 Participants from five countries, and a number of high-level presentations addressed different aspects of the Science of geodetic monitoring of the hydrological cycle. All presentations are available at the Workshop Page, and the key conclusions are summarized in a web-report available at the same page.

The 2nd annual Workshop held from September 30 to October 2, 2010 in Graz, Austria was titled Towards a Roadmap for Future Satellite Gravity Missions. The workshop brought together 55 participants from 11 countries, and it produced considerable output, including a roadmap, several one-page stories demonstrating discoveries based on GRACE, recommendations, and a workshop report. Together with all presentations given at the workshop, the outputs are available at the Workshop page.

With respect to other activities, he reported that in 2008, a number of proposals were submitted, but the limited success indicated a lack of acceptance of geodesy as a valuable tool for hydrology, particularly in the U.S. In 2009, several project-related publications were submitted, and a links with relevant organizations were established.

Hans-Peter Plag identified a number of challenges faced by the project.

Project Objectives

The project objective of precisely monitoring mass movements in the global water cycle has seen some success, particularly based on satellite gravity missions, but the integration of the three pillars was still at a very initial stage. With respect to the detection of reservoir changes, significant discoveries came from satellite gravity measurements, but information flow from science to society is limited. The geodetic observations facilitate global and climate change science and research, but the fact that geodetic data and products are difficult to access and use hampers their utilization in fields others than geodesy. Support for regional water management is still very limited, and the absence of community-validated products is a main obstacle.

Project Goals

Hans-Peter Plag reported that the forum for hydrologist and geodesists was still rather limited with a small number of hydrologist attending relevant geodetic meetings, but very few geodesists participating in hydrological events. The project has contributed to the identification of science and technology challenges, but a general strategy to meet the challenges in developing products for regional water management has not been fully attained. However, the roadmap developed at the 2nd Workshop does specify steps towards future gravity satellite missions. He saw very little progress towards regional products in support of societal application, but found that the understanding of what is needed has improved over time.

Project Approach

Hans-Peter Plag stated that the initiation of new projects was slow and that links to other programs needed improvements. The main success of the project's approach is in organizing the series of annual workshops.

Concerning the next (3^rd) annual workshop, Hans-Peter Plag reported that the proposed title for this workshop was Determination of mass transports in the hydrological cycle from geodetic observations. He initiated a discussion as to the proposed location and time for this workshop, and he asked whether the workshop should be co-located with other events. Further discussion on this was deferred to the end of the meeting.

Concerning new projects, he mentioned that most likely there will be a response to the GEO UIC CfP - and this proposal would be discussed later during the meeting. He also pointed out that other proposals would be crucial for the success of the IGCP Project.

He also emphasized the importance of strengthening links to other programs, including UNESCO's IHP, IGWCO, and the GEO Water Tasks. He also reviewed a list of publications that need to be prepared. In order to make progress towards useful products, he proposed to develop a roadmap for two test regions, namely the U.S. Southwest, and Africa.

Looking forward to 2011, he reminded the participants that the proposed topic of the 4^th Workshop was Integration of geodetic observations and products in models of the hydrological cycle. Because 2011 is rapidly approaching, location, date, and other logistics convcerning the Workshop should be decided soon. He also emphasized the need to present the project at the IUGG meeting in Melbourne, Australia, in 2011. At the latest in 2011, links to regional water management in Africa would have to be established.

The proposed topic for the 5^th, and final, workshop was Improving regional water management in Africa and Asia on the basis of geodetic water cycle monitoring. This workshop should be held in Africa in September-October 2012. The workshop should focus on activities in Southwest U.S. and Africa.

In the discussion, the future of satellite gravity missions was considered. Hans-Peter Plag pointed out that it is still unclear when a GRACE2 mission will be implemented. It is possible and likely that there will not be an immediate follow-on gravity mission. The current research plan of NASA is being driven by the recommendations of the Decadal Study, and the U.S. has made it very clear that no deviations from these recommendations will be accepted. He reported that the roadmap developed during the 2nd Project Workshop was distributed at the GEO-VI meeting in November 2009 in Washington, D.C. Some discussions at GEO-VI indicated that an international trust fund might be a way of getting multi-national funding for a GRACE follow-on mission.

Chris Rizos summarized the Australian contribution to the Project. He reported on activities in environmental geodesy, which were led by Paul Tregoning. A goal of these activities is the validation of the geodetic observations. For the evaluation, the signals of recent droughts are being used. Correlations of different parameters related to land-water storage provide a measure for the signal contents of the geodetic observations. The results of the studies are published in several articles, including Water Resources Research, 45, 2009, Leblan et al., hydrological deformation, GRL, 36, 2009.

The discoveries of rapid ice melting in Antarctica and Greenland were mentioned as relevant for Australia. With respect to GEO and GEOSS, Chris Rizos pointed out that in Australia there is a growing understanding that GEO would need substantial funding for implementation of Earth observation infrastructure.

Peter Fitch presented slides prepared mainly by Albert Vandijk, who leads the Research Team on Environmental Earth Observations at CSIRO (see pdf). He discussed the continental water balance for Australia, and he explain that in Australia, only a small fraction of the total precipitation goes to groundwater recharge, while the main fraction is leaving as Evapotranspiration (about 85%) and runoff (about 10%).

The current ground-based observation system in Australia is sparse, and "in the absence of good information, good decisions are unlikely to occur." Therefore, space-based earth observation has a potentially high value. However, the very low percentage of precipitation recharging the groundwater has kept the priority of GRACE at a low level for many hydrologists in Australia. For the team at CSIRO, the main interest in GRACE stems from the potential to provide constraint for the continental scale water balance model, particularly if GRACE observations are used in a data assimilation framework.

The model developed by the Team at CSIRO aims to combine the best of three worlds: on-ground observations, satellite observations and biophysical models. The approach utilizes multiple constraints and follows the Model parsimony "The simplest theory (model) that still adequately explains all observations." The result is a prototype of the Australian water observation system, which is transitioned to Met bureau for pre-operational testing.

The Team at CSIRO has done some initial work to develop such an assimilation scheme for GRACE observations into the prototype, but currently the Team is not sure of the value and the associated improvements associated of an inclusion of GRACE observations in the model. Before a decision on inclusion of GRACE in an operational system could be made, it would have to be shown that there would be improvements in the predictive skill of the model. Moreover, GRACE would need to have a perspective for continuous missions. However, the modeling work done by the Team could be used to validate GRACE estimates of changes in groundwater storage. Peter Fitch also mentioned that the Australian model could easily be extended to global scale, and for this extension, GRACE would be highly relevant.

The discussion focused mainly on two issues: the reasons for GRACE observations not included in the current prototype, and the high quality of the Australian model. Main reasons hampering the broader use of GRACE data are the lack of suitable validated products and the uncertain future of satellite gravity missions. The Australian model appears to be very well developed and may well be the leading model in the world. In order to carry the discussion between the IGCP 565 Project and the CSRIO Team further, Peter Fitch suggested that Hans-Peter Plag contacted Albert VanDijk. He also made the point that the Australian Bureau might be able to lend some support to those voices arguing for an immediate GRACE follow-on mission. For that, it would be good if CSIRO was able to demonstrate improved predictions using GRACE as a constraint.

Norman Miller reported on a UN Habitat Project titled "Application of Climate Model and Remote Sensed Data to Understand Lake Victoria Hydrologic Variations" (see pdf). The importance of climate change impacts on Lake Victoria Water is underlined by the fact that water is and will continue to be a limiting factor for development in Africa. In order to decrease the vulnerability of water utilities to climate change, quantitative understanding of the risk of climate change is required and can support the planning of adaptation strategies that would increase resilience. In the frame of the project, a Water Evaluation and Planning (WEAP) model has been used on three pilot utilities to quantify the projected water demands and supply limits. The large lake level fluctuations of Lake Victoria cannot be understood without looking into the large-scale forcing of regional rainfall. But the coupling between climate pattern and lake level is currently not well understood. Norman Miller discussed projected changes in the regional patterns and their potential impact on the lake level. The project will create manuals for optimizing water management, but it is not utilizing geodesy.

Norman Miller also mentioned that in a California project focusing on the Central Valley, the approach is rather similar to the Australian approach. A regional model is used to study the driving forces for groundwater changes and GRACE data are being used as constraints.

Chris Hartnady presented the "cGPS for Groundwater Resources Assessment, Hermanus, South Africa" Project (see pdf). Hermanus is located in the Table Mountain Group, which includes the Table Mountain Superaquifer. This large aquifer is a fractured quartzitic aquifer of immense thickness and covers a wide area, which is sensitive to environmental disturbances. Moreover, the area is one of water scarcity, and the level of safe supply has been exceeded several times.

The project utilizes data from the South African TrigNet, which has currently 21 cGPS stations at a spatial spacing of 200 to 300 km. These stations use (recycled) older Ashtech receivers and combines them with new electronics, which poses the problem of getting the old receivers to work with new antennae. The stations also face security issues.

In the Gateway wellfield close to Hermanus, additional cGPS receivers have been installed. Initial results come from 2008-2009 test pumping, which produced well observed signals in some of the GPS time series (both horizontal and vertical components). For the interpretation, 3-D models are required. Anticipated future developments include poroelastic modeling. In terms of observations, the combination of cGPS and gravity monitoring is considered.

Chris Hartnady also mentioned a study on a coastal (meteo) tsunami, which hit the Cape coast in August 2008. Recorded residual wave amplitudes exceeded 0.7 m, and some damage was observed to buildings. There is a slight indication that the cGPS stations at Hermanus recorded a loading signal of this tsunami.

Hans-Peter Plag briefly discussed the status of GPS observations and the available observations of Earth's surface displacements as these relate to hydrology (see pdf). The group at the Nevada Geodetic Laboratory collects and analyzes routinely observations from some 4,000 cGPS sites. However, the spatial distribution exhibits dense clusters in the U.S., Europe, and Japan, while other areas of the world have only sparse coverage. Comparing the spatial distribution of stations to the long-term changes in land water storage as determined from GRACE, he pointed out that many of the regions where significant changes take place are not sufficiently covered with GNSS. Nevertheless, a current project funded by NSF will give an opportunity to study the coherency of hydrological signals in gravity (GRACE) and surface displacement (GNSS) and provide an opportunity to look at spatial scales between 10-300 km. A key challenge is in the different spatial resolution of the available observations of gravity changes and surface displacement. Norman Miller added that this study could be done jointly with the California project.

Chris Hartnady presented the proposal "Earth Observation Techniques - Global Navigation Satellite System (GNSS) and Interferometric Synthetic Aperture Radar (InSAR) - for Groundwater Water Resource and Drought Management" (see pdf), which had been submitted in response to the first round of the GEO UIC Call for Proposals (CfPs), and it was accepted with some concerns. He reviewed the main goals of the CfP and emphasized the need to link the project to decision makers. The goals of the proposal were fully in overlap with the IGCP 565 Project. Participation was mainly recruited from IGCP 565 Project members. The main focus was to develop integrated ways of using GNSS, InSAR, and GRACE for studies of deep groundwater reservoirs.

It is anticipated that the full proposal will be submitted some time in early 2010. Hans-Peter Plag reported that the GEO UIC and CBC were still in discussion about the conditions of the CfP for full proposals. In the discussion, it was pointed out that the project should include a significant capacity building effort. The participants expressed their support for the submission of a full proposal.

Eric Calais reported on the AfricaArray-Proposal, which was submitted to NSF (see pdf. Currently, AfricaArray is a predominantly seismic network. The proposal aims to augment the network with additional sensors serving a wide range of geoscience applications. The proposal asks for a number of new cGPS stations, which would serve applications in geodesy, hydrology, atmospheric water vapor contents, and space weather. Low-latency products are already being used for Numerical Weather Prediction in South Africa. The proposal was under review, and a proposal for a workshop to develop the science application was submitted.

Eric Calais stressed that there is considerable ground-based infrastructure available but data are either released after long delays or not at all. Therefore, science P.I.s should be urged to release data with low latency. These science P.I. benefit from the global community making thousands of stations available, and it would be reasonable to expect that they give something back to this community.

A discussion arose to what be the best theme for the AfricaArray workshop. The decision was made that a regional approach should be considered, with a clustering of the infrastructure instead of aiming for a thin (too thin?) cover of all Africa. One suggestion was that west Africa could be a region for hydrology.

Hans-Peter Plag suggested a Syndrome approach, where the different geoscience applications for the interesting regions are plotted. By overlaying these maps, intersecting hot spots with many applications for the infrastructure could be identified.

It was also mentioned that it would be important to build up from existing cores and areas of strength. Additional funding might come from the Global Facility for Disaster Risk (World Bank) for hydrology. Here, floods could be a focus.

Richard Wonnecott discussed the status of AFREF (see pdf). Currently, 57 stations are collecting data for AFREF, however, 22 of these stations are not working or providing data. Eleven of those stations delivering data are in South Africa. He informed that the UN Economic Commission for Africa is funding a workshop on the replacement of GPS stations in Feb. 11-12, 2010, in Nigeria.

Julia Peffer presented the project "The GHYRAF (Gravity & HYdRologyin AFrica) experiment" and reported first results on land water storage changes (see pdf). GHYRAF is a pilot study investigating the water cycle in Africa from the Sahara to the equatorial monsoon zone with ground and satellite-derived observations of gravity changes and surface displacements. The goal is the comparison of hydrology models and a range of observations including gravity, surface displacement, hydrology, meteorology, etc. The project combines episodic gravity measurements at reference sites, continuous observations of surface displacements, micro-gravity surveys, and hydrological observations with larger scale GRACE measurements. The project team includes a number of French and Luxembourg science institution as well as local water authorities in Africa. Julia Peffer described the observations campaigns, reviewed the observations and discussed initial results of modeling attempts. To a large extent, models explain the observed signal in terms of land water storage. A range of project activities is planned for 2010 including several measurement campaigns.

In summary, the first cycle of absolute gravity measurements demonstrates the feasibility of the experiment with low noise observations despite the variability of the climate, a lack of electric power and other logistic problems. The preliminary processing of GPS vertical motion revealed clear hydrological signals related to monsoons. The observed in situ gravity changes could also be related to hydrological signals. The GRACE solutions were found to agree well with hydrology models for the Niger and Chad basins.

In the discussion, the participants showed great interest in the projects, and the wish to have a link between the IGCP 565 Project and GHRAF was emphasized. Hans-Peter Plag was asked to discuss further links with Jacques Hinderer.

The scope of the third annual Workshop to be held in 2010 was discussed. There was consensus that the separation of hydrology and tectonics in geodetic observations is still a major challenge, and therefore, the workshop should focus on this problem. It was agreed that the workshop should take place in the Southwestern U.S. where large hydrological and tectonic signals superimpose in geodetic observations. The workshop should also look into other regions, including East Africa and some areas in Asia.

Norman Miller volunteered the University of Berkeley as a potential site for the workshop. Proposed members in the Program committee included Eric Calais, Jay Famiglietti, Chris Hartnady, Norman Miller, and Hans-Peter Plag. The proposed time window was September to October 2010.

The plan for the 2011 annual workshop was also discussed and further detailed. Since this workshop will focus on assimilation of geodetic observations in hydrological models, it was suggested to hold this workshop in Australia in coordination with CSIRO and co-locate it with the IUGG meeting. In Australia, the state-of-the-art in hydrological modeling is rather advanced. At the IUGG meeting, a symposium on geodesy and hydrology will be jointly organized by IAG and the International Association of Hydrological Sciences. Thus, organizing the workshop in Australia would benefit from both the hydrological modeling and the symposium.

The meeting provided an analysis of the project progress and achievements and identified the challenges for the next three project years. The summary of the outcomes of the first two IGCP 565 Workshops illustrated the potential of the geodetic observations for hydrological applications, and it emphasized the importance of continued satellite gravity missions for monitoring the global water cycle on regional to global scales. Challenges for the project include providing a platform for a better communication of hydrological and geodetic communities, and a better integration of the project into the frame of other relevant programs, such as GEO, IGWCO, GEWEX, and UNESCO/IHP. Utilization of geodetic products for hydrological applications is hampered by a lack of well validated and quality controlled geodetic products easily applied by hydrologists, and a challenge for the project is to make these products available to those who can put them to best use.

An important results was an improved description for the scope of the next two annual workshops. A key issue in making full use of geodetic observations for the detection of land water storage changes is the separation of the hydrological signals from other geophysical signals in the geodetic observations. In many regions, tectonic signals have similar spatial and temporal scales as hydrological signals. Therefore, the third workshop will focus on the separation of tectonic and hydrological signals in geodetic observations. Assimilation of geodetic observations in hydrological models is considered the most promising avenue for utilization of these observations for regional water management. The 2011 annual workshop will focus on assimilation.