CPTEC/INPE

Cachoeira Paulista – SP, Brazil

4 – 6 December 2000

Introductions and General Framework Presentations

Charles Vörösmarty and Jose Marengo welcomed participants on behalf of the two host institutions, CPTEC and the University of New Hampshire, and gave a preview of the agenda for the three-day workshop: one day of research presentations, and two days of practical use of the HydroNET system, including hands-on instruction and working group discussions.

Marengo requested that everyone give short presentations that would emphasize the significance of their research within the context of LBA and the Amazon region’s hydrometeorology.

Nobre

Carlos Nobre welcomed the group and gave a brief overview of the LBA program, its current status, and CPTEC’s role.

Most of the workshop participants are active within LBA, which began as an ambitious idea: an all-encompassing experiment across many disciplines for understanding the Amazon region’s natural ecosystems and how humans interact with the natural setting (biologically, physically, etc.) in one of the world’s most important biomes. Similar projects are now beginning in Southeast Asia and Africa. This workshop clearly depicts the concept of Earth Systems Science and multi-disciplinary investigations, bringing together a mix of hydrologists, ecosystem modelers, ecologists, and hydrometeorologists. Such collaboration puts LBA a step forward in science.

LBA had been planned for many years, but after research started in 1998, the program "took off" and is now producing significant results. More important experiments are still to come. As there is much work left, this workshop plays a key role in highlighting hydrometeorological questions and what we need to do in order to accomplish our goals. In Rondônia, for example, we still have much to learn, and are seeing ongoing research building on knowledge being gained. There have already been decades of hydrometeorological data collection in South America, but typically not very much interpretation of that data. It is time to begin extracting a deeper scientific understanding based on these data.

The new INPE/CPTEC center was inaugurated in 1994. It has grown alongside LBA, playing an important role in its planning and implementation. It also houses the LBA Central Office and contributes much to the program’s modeling efforts, especially meteorological modeling, by running a mesoscale numerical weather prediction model and housing one of the data systems. CPTEC provides a test bed for a lot of new knowledge being gained.

There remains an identifiable gap: from the inception of LBA, there should have been more studies in mesoscale hydrology. There is a broad array of studies, especially in Rondônia and Santarém, emphasizing other key research areas (e.g., land-atmosphere exchanges, ecology), but we have never been able to reach the same level regarding hydrological process studies. LBA has focused more on large-scale issues using remote sensing, but it might be beneficial for this community to discuss that at some point in the workshop and consider what is being missed. This workshop is a very worthwhile opportunity; perhaps we should have held it several years ago.

Knapp

David Knapp offered some logistical information concerning the morning’s schedule, lunch plans, computers available (17, each with a CD-ROM reader and internet access), restaurant suggestions, and his own contact information. He added that any printouts that would need to be made during the workshop could be done through him.

White

Ellen White gave an overview presentation of the LBA-Hydrometeorology Project, defining its context as a NASA LSHP-funded program within GEWEX, of the World Climate Research Program, as well as within the broader LBA experiment. She then briefly summarized the eleven LBA-Hydrometeorology investigations, each of which represents a joint collaboration between Brazilian and US investigator teams. The investigations are as follows:

Vörösmarty

Vörösmarty spoke concerning the scope of the workshop, expressing eagerness to see how the workshop would facilitate collaboration among regional partners sharing hydrometeorological data, and recognizing the various sources of funding support for the workshop. The majority of support came from NASA’s Land Surface Hydrology Program (LSHP), but with support from UNESCO, CATHALAC, and IGBP as well.

Overall goals of the LBA-HydroNET project are (a) to construct a GIS-based Web site to serve as a repository of hydrometeorological information for use by LBA researchers; and (b) to develop a basin-scale water budget closure system for Amazonia using observations and simulation outputs. Vörösmarty and CPTEC colleagues have been especially interested in seeing this happen for several years.

There is great concern about the global phenomenon of decline in hydrometeorological monitoring capability due to loss of ground stations in what used to be well-monitored parts of the Earth. A time series of this decline has been compiled and documented by the World Meteorological Organization’s Global Runoff Data Center. The number of reporting stations shows an exponential increase until the mid-1980s, then a sharp decline. The apparent reason for this was the habitual delay in making data available from the time of collection, loss of networks, and strict restrictions on data availability and use. The UNH group has been active in working against this trend on a global scale. For the Arctic region, they gathered colleagues from Canada and USGS, pooled existing archives, mounted data in easy-to-use formats, and multiplied usable data by a factor of 25, to approximately 3700 active meteorological stations.

R-HydroNET—the continental-scale counterpart to LBA-HydroNET—was begun three years ago. Original funding from the U.S. State Department and UNESCO was secured in summer 1997, the Web site was established and operational by spring 1998, and a workshop for R-HydroNET was held in Panama City in summer 1998. In that same year, the LBA-HydroNET proposal was submitted and accepted for funding under the NASA research announcement. At this point, the LBA-HydroNET v1.0 framework has been created, and this workshop marks the beginning of the next phase of project development.

Some decisions will need to be made concerning the direction of the organization and data management of LBA-HydroNET. Should data be limited in spatial scale to Amazonia alone, or include more/all of South America? In order to meet needs and desires of data providers and users, data is currently categorized in a tiered fashion, with Level 1 being publicly accessible archives, Level 2 having password protection, and Level 3 consisting of completely restricted data sets. (Vörösmarty showed examples of the Web site’s data presentation, what is included, how it is organized, and how value-added data products are incorporated.) Will we want to continue this "security" system, or make changes, realizing as well the general data policy guidelines of LBA more generally? The time domain is currently 1980 to present, and the spatial domain includes the full Amazon Basin.

The workshop should concentrate on specific products and activities. By considering concrete examples of important variables, it becomes easier to determine what would be the preferred spatial resolutions, temporal resolutions, and data sources. Potentially, using the HydroNET system, a researcher could assemble data sets and compare them to see where they do and do not agree in order to improve the accuracy of data and value-added products, to learn how to debug programs more effectively, etc. This workshop may be a good opportunity to brainstorm other applications. The team working on the water budget closure system, for example, is seeking to identify several conceptual domains for viewing key water balance elements. Using the HydroNET system, investigators could take various data sets, plug them into equations, and quickly evaluate the quality of the data sets. Often problems with data can be perceived only in combination with other data.

The goals of this workshop are:

• to announce the creation of the Web-based data bank (LBA-HydroNET v1.0)
• to demonstrate the system’s current capabilities
• to create an initial steering committee composed of regional data partners who will assist in design and implementation of forthcoming versions of the LBA-HydroNET, develop research and applied scientific applications for the system, and lead future regional data gathering, assimilation, and production efforts using the system
• to provide initial training in the use of the HydroNET
• to test the system using regional hydrometeorological data: import a first set of input data provided by LBA-HydroNET partners, process into value-added hydromet products, and mount on WWW site for use by data partners.

Marengo

Marengo spoke briefly about applications of the HydroNET system over the LBA domain. In Amazonia we have a place where we can carry out our research, and there are several neighboring regions that are closely linked in terms of hydrometeorological processes, events, and their effects. If we can map precipitation in an area and avoid flooding catastrophes, clearly this should be a priority. One of the requirements of LBA is the free availability of data. Conceived as such, this project can have major impacts. A long vision is clearly necessary and possible.

Hurtt

George Hurtt presented an overview of LBA-Ecology and its scientific relationship to LBA-Hydrometeorology. Most ecological modeling proceeds by assuming that atmospheric and hydrological data can be specified as exogenous inputs. Hydrologists seem to take the opposite tack, using biospheric measurements as inputs for land-atmosphere models. It is, however, more appropriate to consider it as a coupled system.

Important progress can be and has been made by studying these systems in isolation. Research must be done in this way to understand the component systems: their responses, mechanisms, scaling issues, etc. The central LBA questions, however, are of the coupled system, which is important because of potential feedbacks across the components.

Feedbacks can dramatically amplify responses (e.g., water vapor feedback on a climate system), and they can fundamentally alter systems. Land use change, for example, initiates a process where temperature and precipitation may be altered, which will in turn affect natural vegetation, potentially to an even greater degree than the initial land use change that affected it.

LBA-Ecology and Hydrometeorology need to work together. The practical benefits are that Hydromet gets relevant details about land surface, and Ecology gets weather, climate, and hydrological information. The "big picture" strategy of these two groups working together should be to move toward coupled models that would simultaneously conserve carbon, water, energy, and nutrients, and to develop the ability to identify, assess, and better predict feedbacks between systems.

Smith

Eric Smith briefed the group on another potential front for collaboration and data sharing, the Global Precipitation Mission (GPM), an endeavor planned by NASA on which he will serve as Mission Scientist.

Official participants of the GPM program currently include the U.S. (NASA) and Japan (NASDA), with additional participation expected from Italy and possibly a number of other countries, including China, Brazil, Canada, and Korea, as well as sister agencies such as NOAA. The mission is intended to build on the work of TRMM, involving a core satellite carrying active and passive instruments as well as nine member satellites with passive sensors. The goal is to produce a nearly global (within the 60 degree latitude) precipitation measurement with a sampling interval of 3 hours or less (the sampling frequency given as a minimum by hydrologists). There will be a need for very good validation capabilities, which is a key point of potential participation by LBA investigations.

The targeted year for commencement of the mission is 2007, viewed as the first phase of an ongoing series of satellite launches, building into an international-scale venture. Hardware should be selected by 2003, and the satellite constellation should be operational by 2007, but some instruments are expected to fail and/or wear out, etc., as time continues.

GPM is poised to become a centerpiece for Global Water Cycle science. The advanced dual frequency radar system of NASDA-CRL will lead to improved rain retrievals through the capacity of its measurements to be sensitive to precipitation DSD variability, and improved precipitation data is extremely valuable for many areas of research, particularly crossover areas of interdisciplinary interest. GPM offers a program for blending scientific research and applications, which can enable the advance of space technology and the attainment of new levels of international cooperation in science. Using flexible mission architecture and international partnering, GPM presents a path for the development of an operational and space-based global rainfall measuring system.

Individual Participant Presentations:

Vörösmarty welcomed the group back, asked for submission of missing abstracts, data for uploading into the system, and presentation slides for inclusion in the report of the workshop. He suggested that breakout groups for the following day be defined according to applications of the HydroNET system, and gave some thoughts on issues that would need to be addressed:

• precipitation, and related, mutually constraining parameters like runoff
• data policy (raw data vs. value-added products, delays in mounting of data in order to protect products provided by operational science partners)
• data rescue (potential need to digitize tabular data and/or maps)
• coupled water, energy, carbon models
• real-time vs. time-series analysis
• links to other LBA activities (fire, LC/LUC, C budgeting)
• use of runoff to "condition" precipitation
• regions of interest, like Andes or Amazon floodplain
• extreme events/ENSO

Discussion

Gentry: Beija-flor demonstration

Gentry began by describing the Metadata Editor, explained harvesting and the reasons for simultaneous harvest at ORNL and CPTEC. She emphasized that if a provider stipulates in the data set that he/she does not want it to be harvested, it will not appear in the index.

In the event that metadata is entered incorrectly, the provider can enter and fix it, whether it is located at ORNL or at CPTEC, or the coordinator at that site can do it. If someone else notices the error, the best thing to do is to contact either the data contact or the provider.

Gentry showed a slide of the Beija-flor search page, which is currently in English with a draft version in Portuguese, as is also the case with the user’s guide. Files are in XML, and users can link to other search engines from within Beija-flor. Also, if IAI-DIS, ESIPs, or other members of the data archiving community wish to submit metadata files, the capability exists for searching those indexes as well. Gentry explained how to perform searches, including using a map for spatial searches. She listed data sets that are currently held in the archive at ORNL, and concluded by demonstrating the processes of performing a search and entering metadata, noting that soon the Metadata Editor will include two file type options: data set and poster (where presently the only option is "metadata").

Vicente: Can data files be submitted in different formats?

Schloss: EOS-Webster demonstration

EOS-Webster is a participant member of the ESIP/LBA Cluster. It is a digital library that provides model products and GOES and IKONOS products, as well as services such as creating spatial and temporal subsets and reformatting data, offering the user a choice of media. Spatially explicit model data include ED, NOAA/NASA Pathfinder, and Amazonian vegetation and soil maps from WHRC. Schloss demonstrated how to access data documentation, showed the page listing data collections, and showed the search and retrieval tool.

Fekete/D'Almeida: LBA-HydroNET system demonstration

Fekete and D'Almeida led the group in a demonstration and hands-on practice session with the LBA-HydroNET system in the CPTEC computer lab.

Discussion: Working Group designations

Four groups were formed, around

1. precipitation issues
2. coupled models
3. water budget closure
4. operational hydrometeorology.

Groups were to meet independently the following morning, to discuss key scientific questions, operational objectives, data needs, important tools to be provided by LBA-HydroNET, and "governance" issues.

Guidelines for Break-out Working Groups:

I. Scientific Questions and Operational Objectives of LBA-HydroNET

1. Identify key scientific questions that will be addressed by the sub-group.
2. Identify:

II. Data Needs

1. Perform initial data inventory

1. What data are already included in LBA-HydroNET and ORNL data lists relevant to this area of study? (Consult hand-outs from Tuesday.)

2. What are the principal data gaps remaining and which data sets are recommended for collection under LBA-HydroNET?

1. Are there paper products such as hydrographic yearbooks or maps that require digitizing? (If so, identify these data explicitly.)

2. Domain of the data sets

1. Spatial: all of South America, everything north of 40° S, only Amazonia?

2. Time: Start: 1960s, 70s, 80s (??). End: 2000 and beyond.

3. Resolution of existing or recommended new data sets

III. What tools should LBA-HydroNET provide users to support these studies?

Examples include:

• Visualization for rapid error identification
• GIS-based sub-setting
• Tools for rapid inter-comparisons
• Others, please specify

IV. "Governance" (how to manage the system)

A. What strategies or mechanisms can be developed to:

• Respect the interests of data providers, while also
• Ensuring the delivery of timely and freely available LBA science data sets?
• How should raw data be treated?
• How should value-added products be treated?

1. What linkages should be made between LBA-HydroNET and:

Two volunteers from each group became working group coordinators, to communicate expectations, to track post-workshop progress toward goals, and to serve as an ad-hoc steering committee.

Working group reports: