In mid-January 2014, results of phase 6 of the Global Earth Observation System of Systems (GEOSS) Architecture Implementation Pilot (AIP) were presented at the 10th Summit of the Group on Earth Observation in Geneva. Author, Dr. David Arctur managed a yearlong project within AIP-6, to advance the Water Societal Benefit Area (SBA) of GEOSS. This article discusses the results of the project.
Water resources are essential to all life on Earth, and the sharing of water data is essential if we are to make the best use of available resources and get better at understanding and preparing for floods and droughts. While the developed countries have made the most progress in this direction, the importance of building and nurturing Earth observation and analysis skills and tools among developing countries cannot be overstated; this is a major “digital divide” and barrier to international cooperation on environmental sustainability issues. Formed in 2002, the Group on Earth Observation (GEO) has been working to overcome the institutional and interdisciplinary challenges to publishing, finding and accessing environmental data across many subject domains.
In mid-January 2014, results of phase 6 of the Global Earth Observation System of Systems (GEOSS) Architecture Implementation Pilot (AIP) were presented at the 10th Summit of the Group on Earth Observation in Geneva. For more on GEO and GEOSS, see http://earthobservations.org and http://geoportal.org. The GEO X Summit included a Ministerial Summit on the last day, attended by representatives from the administrations of the 90+ member countries, who voted unanimously to continue the GEOSS initiative for another 10 years.
My part in this was to manage a yearlong project within AIP-6, to advance the Water Societal Benefit Area (SBA) of GEOSS. Other teams in AIP-6 worked on Energy, Agriculture, Biodiversity, and Disaster Management SBA projects. All of this is to improve the accessibility and application of Earth observation data and models globally, with particular attention to “capacity building” among developing nations. See http://www.ogcnetwork.net/aip6responses for the complete set of work proposals accepted for AIP-6; http://www.ogcnetwork.net/system/files/AIP6_GEOSS_Water_Services_Proposal_final.pdf for our team’s proposal.
Our goals for the AIP-6 Water SBA were: (1) to improve discovery and access to water resources data around the world; (2) to improve integration of gridded and time series data; and (3) to improve the tools and processes for federating the regional and national picture to a global system (GEOSS), while demonstrating the benefits of an already operational federated regional water data within a country. We did all this, and I think we’re onto something.
Part of what made this work was an incredible team. The science lead was Professor David Maidment from the University of Texas at Austin, where he is Associate Director of the Center for Integrated Earth System Science (CIESS). He was able to call together colleagues around the world and say, “let’s do this.” I work with Dr. Maidment at UT CIESS, kept the objectives in front of us, and coordinated the team efforts and final reports. We also had Jim Nelson and Dan Ames from Civil and Environmental Engineering at Brigham Young University (BYU). Jim was networking with water management agencies in Latin American countries. Dan’s a key developer for the Consortium of Universities for the Advancement of Hydrologic Science (CUAHSI) HydroServer and HydroDesktop software tools. We had Alva Couch, director of the CUAHSI Water Data Center, recently moved from UC San Diego to Tufts University in Medford Massachusetts. From the Global Institute for Water Security (GIWS) at the University of Saskatchewan in Canada, we had Branko Zdravkovic pulling together Environment Canada’s precipitation monitoring network of 829 gauge stations.
For the national and regional water agencies, we had Martina Bussettini from the Italian National Institute for Environmental Protection and Research (ISPRA), and Silvano Pecora from the Regional Environmental Protection Agency (ARPA-ER) of Emilia-Romagna region in northern Italy, one of Italy’s 21 water regions. We had Jochen Schmidt from New Zealand’s National Institute for Water and Atmospheric Research (NIWA). And we had two technical staff, Sean Hodges and Brent Watson from the Horizons Regional Council, one of New Zealand’s 16 regional water-reporting agencies.
Representing the Community on Earth Observation Satellites and adapting the CEOS Water Portal for this project were Satoko Miura and Yoshiyuki Kudo from JAXA (Japanese Aerospace Exploration Agency). The European Commission JRC (Joint Research Centre in Italy) and ECMWF (European Centre for Midrange Weather Forecasting in UK) together worked on producing WaterML data services for flood forecast products with GloFAS (Global Flood Awareness System). Representing the software industry, we had folks from Esri based in the US, Kisters AG from Germany, and PYXIS Innovation from Canada. And finally, we had critical and responsive engagement with designers and developers of the GEOSS architecture, particularly at the Italian Research Council Institute of Atmospheric Pollution Research (CNR-IIA) and at George Mason University in the US.
Water Data and Maps
The GEOSS Portal provides metadata and access for millions of datasets, but it is not necessarily the best portal for interacting with the datasets you find, as it is intended mainly for identification and download of whole datasets and links to other relevant portals. What we wanted to do with this project was to provide a means for users to discover and browse maps of water data, with the ability to see and download time series of streamflow and precipitation, site by site. For this we used the Esri ArcGIS Online web map viewer, the CUAHSI HydroDesktop map client, and PYXIS WorldView. We served maps of locations and basic time series descriptions of stream gauges and precipitation monitors, using these products' OGC WFS (Web Feature Service) compliant interfaces. Also, we provided links in certain data fields about each gauge to supply three main types of data: (1) a graph of current or historical streamflow and precipitation time-series values; (2) a data file in XML format using the OGC WaterML 2 (or similar) schema corresponding to the same data as in the graph; and (3) a comma-separated-value (CSV) file with the same data values, that could be read by Microsoft Excel or another spreadsheet program. Figure 1 shows the map of worldwide streamflow gauges, and Figures 2 & 3 show data content for a stream gauge in northern Italy.
Figure 1. Map of global streamflow gauges
Figure 2. Stream gauge data in Italy
Figure 3. Stream flow graph for Po a Boretto, Italy
A very similar procedure brings up precipitation gauges around the world, and shows data for specific gauges, see Figures 4, 5, 6. Working with Canada’s GIWS, we have mapped 829 precipitation monitors operated by Environment Canada (red dots in Figures 4 & 5).
Figure 4. Precipitation monitor stations map
Figure 5. Loon Lake monitor in Saskatchewan
If you look closely at the Loon Lake monitor description panel in Figure 5, you’ll see that the data field labeled “EndDate” is blank (empty). This is a convention we adopted to mean that the gauge is collecting current, real-time data (i.e., no end date), to distinguish such gauges from the many gauge sources that just report data collected over a fixed historical period.
Figure 6. Precipitation values for Loon Lake gauge
With completion of this project, there are now several countries using WaterML for time series data, and web mapping for gauge locations. So this is how we’ve addressed the first objective I mentioned near the beginning of this article.
Integration of Gridded and Time Series Water Data
The second objective was to improve integration of gridded and time series water data. The water variable we chose for this was soil moisture, specifically the top one-meter of soil moisture, which is a model output from NASA’s Global Land Data Assimilation Program (GLDAS), based on satellite data. This program generates grids of soil moisture and other water data over all the world’s landmasses, at one-quarter-degree spacing, with 3-hour intervals, from 1979 to the present (and ongoing). Working with Esri and Kisters, we have spurred development of a soil moisture map that can be tied to the time series at any grid point, see Figure 7.
Figure 7. Soil moisture map with time series from Tunisia
This is a web application that allows the user to pick any point on a landmass, and it will reveal the time series of values at the nearest grid point in the lower frame. The overall map can be animated to step through a sequence of global soil moisture maps for a given time range. The user can interact with the graph and the time line at bottom, to pick the global soil moisture time step to be displayed. In addition to mapping and graphing soil moisture through time, this application allows comparison of multiple locations, and multiple time ranges at a single location, by automatically overlaying and scaling multiple graphs, see Figure 8.
Given that the NASA LDAS model outputs are not equally accurate over the entire globe, this capability will help in comparing the model outputs with in situ ground-based observation stations. This will also help us better understand soil moisture where it has not been measured.
Figure 8. Soil moisture map comparing time series at two locations
Federation of Regional and National Water Data
The third objective of this project was to improve the tools and processes for federating regional and national water data. We started this project hoping to stimulate federation of national sources into a global picture, and we have an enthusiastic start on that with the countries represented. We found that many countries have several regional Hydrological Services, such as in Italy (21 regions) and New Zealand (16 regions). The national water agencies in these countries depend on all their regions to submit accurate, timely, and consistent reports of water resources information in order to develop a national picture of their overall situation. In practice, this has been quite difficult and time consuming, with the result that such a national picture is never available “on demand” but only at predetermined reporting intervals.
As a result of this project, and thanks to the motivated efforts of participating regional water managers in Italy and New Zealand, these countries are now much closer to being able to generate on-demand national water status reports. In Italy, the example set by the ARPA-ER regional manager has led to publishing a consistent set of water maps for all 21 regions. In New Zealand, so far two regions and two national institutes have implemented the data mapping services developed there during the last year. This was a very successful start and provides a model for other New Zealand agencies to follow. To support the wider adoption of water data services in New Zealand, several agencies are currently working on a New Zealand environmental time series exchange standard based on OGC services. In Latin America, outreach from Brigham Young University in Utah has brought the water agencies of Mexico and Dominican Republic to publish selected water data. Nicaragua, Honduras and Guatemala are now starting to get up to speed on the technology, mainly held back by political and national security concerns. We hope to see further progress with these and other Latin American countries in future projects.
What’s next for AIP-7
The next cycle of GEOSS implementation in 2014 will start soon. We plan to focus on bringing additional authoritative (WMO) sources of water data into web maps from the Global Precipitation and Climatology Center (GPCC) and the Global Runoff Data Centre (GRDC), both in Germany. We’d also like to develop maps for water quality data, which is under the auspices of the United Nations Environment Programme (UNEP). A recent proposal to OGC for representing water quality data in WaterML as a community Best Practice could be very timely for this effort. We also hope to add more countries to the list of those supporting WaterML data services, and are expanding our outreach to developing countries in Latin America through access to CUAHSI resources in “The Cloud”. It should be another good year.