Appendix B
Planning, Designing, Operating, and Utilizing the Results from an Integrated Observational-Modeling System

The planning, design, operation, and utilization of an integrated observational-modeling system involves many elements, or stages. Some of these are scientific or technological, whereas others are organizational or social. Eight such elements are summarized here.

These include (1) defining goals, which may include specific “deliverables” for a narrowly defined research project or flexible targets when the project is established for broader and potentially changing uses; (2) building an initial team with appropriate expertise to define and oversee accomplishment of the goals, but often allowing the team to change over time as a project evolves; (3) designing the project to achieve the goals, either specifically or with flexibility to allow for multiple-use data; (4) collecting and validating the data, integrating and validating new data collection methods as appropriate over time; (5) organizing the large data sets for a variety of different uses; (6) integrating observations across sensors and networks; (7) merging the integrated observations with models and model validation; and (8) delivering the information products from the integrated observations and merged observation-modeling to those applying them to flood and drought forecasting, water management planning, disaster response, source water protection, and other areas.

These eight elements often must be addressed in an iterative fashion as a project evolves. For example, once information is delivered, managers or other members of the scientific community may suggest changes to an ongoing project to meet additional or changing needs. The elements are discussed briefly below and are either explicitly or implicitly part of the studies summarized in Chapter 4. As the case studies were initiated for different reasons and are ongoing, each shines a light on different elements or sets of elements.

  1. Defining project goals and “deliverables” is, of course, part of any pro-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 187
Appendix B Planning, Designing, Operating, and Utilizing the Results from an Integrated Observational-Modeling System The planning, design, operation, and utilization of an integrated observa- tional-modeling system involves many elements, or stages. Some of these are scientific or technological, whereas others are organizational or social. Eight such elements are summarized here. These include (1) defining goals, which may include specific “deliverables” for a narrowly defined research project or flexible targets when the project is established for broader and potentially changing uses; (2) building an initial team with appropriate expertise to define and oversee accomplishment of the goals, but often allowing the team to change over time as a project evolves; (3) designing the project to achieve the goals, either specifically or with flexibility to allow for multiple-use data; (4) collecting and validating the data, integrating and validating new data collection methods as appropriate over time; (5) orga- nizing the large data sets for a variety of different uses; (6) integrating observa- tions across sensors and networks; (7) merging the integrated observations with models and model validation; and (8) delivering the information products from the integrated observations and merged observation-modeling to those applying them to flood and drought forecasting, water management planning, disaster response, source water protection, and other areas. These eight elements often must be addressed in an iterative fashion as a project evolves. For example, once information is delivered, managers or other members of the scientific community may suggest changes to an ongoing pro- ject to meet additional or changing needs. The elements are discussed briefly below and are either explicitly or implicitly part of the studies summarized in Chapter 4. As the case studies were initiated for different reasons and are ongo- ing, each shines a light on different elements or sets of elements. (1) Defining project goals and “deliverables” is, of course, part of any pro- 187

OCR for page 187
188 Appendix B posal process. In some cases, these goals appear to be deceptively straightfor- ward. For example, for the Comprehensive Everglades Restoration Plan (see Chapter 4), “The overarching objective of the Plan is the restoration, preserva- tion, and protection of the South Florida ecosystem while providing for other water-related needs of the region, including water supply and flood protection” (Water Resources Development Act of 2000). However, enormous amounts of time and energy have been—and continue to be—invested to define what “resto- ration” constitutes, and what the end points might be. Major multidisciplinary scientific initiatives wrestle with integrating multiple objectives, such as under- standing fundamental processes such as streamflow generation, investigating scaling relationships of observations over time and space, understanding behav- ior under extreme conditions, and developing new instrumentation. The more multidisciplinary the project the more difficult—and more critical—it is to es- tablish one’s goals at the onset. In many cases, it is essential to allow the goals to change over time as new methods are developed, new ideas evolve, and new researchers add to both the needs and the capabilities of the project. (2) Building a strong, interdisciplinary team, when the project spans disci- plines, is as essential as it is supremely challenging. As one participant in an NRC workshop expressed it, “A ‘multidisciplinary’ team is put together and they work in isolation until the very end, when they fight.” Some of the difficul- ties are neither scientific, nor institutional, but personal. Keys to success in put- ting together an interdisciplinary group include finding colleagues who work at institutions that have policies and practices that lower barriers to interdiscipli- nary scholarship, and are willing to “immerse themselves in the languages, cul- tures, and knowledge of their collaborators” (NRC, 2004). (3) Designing a project to achieve the goals set out, whether narrow or broad, specific or flexible, is the next step. An overall approach for the particu- lar needs of interdisciplinary collaborations is described in Benda et al. (2002) as follows: [T]he success of interdisciplinary collaborations among sci- entists can be increased by adopting a formal methodology that considers the structure of knowledge in cooperating disci- plines. For our purposes, the structure of knowledge com- prises five categories of information: (1) disciplinary history and attendant forms of available scientific knowledge; (2) spa- tial and temporal scales at which that knowledge applies; (3) precision (i.e., qualitative versus quantitative nature of under- standing across different scales); (4) accuracy of predictions; and (5) availability of data to construct, calibrate, and test pre- dictive models. By definition, therefore, evaluating a structure of knowledge reveals limitations in scientific understanding, such as what knowledge is lacking or what temporal or spatial scale mismatches exist among disciplines.

OCR for page 187
Appendix B 189 This process, if followed at the project formulation stage, can be used to construct “solvable problems”, and involves building consensus among team members with respect to precision requirements, scales of analysis, and discipli- nary expertise needed. A further advantage is that it leads to a feedback loop to examine whether the project goals and “deliverables” as originally conceived may need to be modified or rejected (Benda et al., 2002). (4) Collecting and validating data as it relates to sensors is covered in Chapter 2. Other aspects of data collection and validation, while central to inte- grated observing systems, is beyond the scope of this report. This is clearly an enormous field by itself. The Environmental Protection Agency’s Field Opera- tions Manual for Wadeable Streams (EPA, 1998) details protocols for a wide variety of activities from stream discharge measurements to periphyton sam- pling. The U.S. Geological Survey lists method, sampling, and analytical proto- cols for a variety of dissolved solutes and suspended materials at http://water.usgs. gov/nawqa/protocols/ methodprotocols.html, and biological sampling, habitat, and laboratory protocols at http://water.usgs.gov/nawqa/protocols/bioprotocols.html. (5) The organization of data-sets is another important step. It is dealt with in this report in the context of smart sensors and sensor networks, which can help to avoid the collection of large amounts of time series data at times when little change is occurring in the measured parameter, and to reduce data where this is deemed useful. With satellites, the amount of data generated can be enormous—on the order of a terabyte per day for a single satellite (NRC, 2000a) and thus overall several thousand terabytes per year. In fact, in many cases the size of the data archives is growing faster than we can derive information from them; NASA’s Earth science data holdings increased by a factor of six between 1994 and 1999 and then doubled again from 1999 to 2000 (Climate Change Sci- ence Program, 2003). However, this is also a highly parameter-specific activity, and it is difficult to generalize principles without a specific context. (6) Integrating observations across sensors and networks: Currently vast amounts of environmental and water-related information are collected daily by a wide range of sensors, and these data are being used widely for water manage- ment, water-quality monitoring, flood hazard forecasting, and so forth. Exam- ples of such applications are provided in the case studies in Chapter 4. Sensors range from snow measurements taken manually by the National Weather Ser- vice to experimental embedded network sensors to control storm water dis- charge. And in between, there is a tremendous range of operational and experi- mental sensor platforms or stations that collect, store, and transmit data in a va- riety of ways. Currently, most sensor platforms are unable to communicate di- rectly with each other, and there is a lack of inter-operability among data net- works, for the most part, which will be discussed below. New developments in sensor technology are occurring rapidly, both in the ability to obtain measurements in new novel ways (for example, through biosen- sors for water quality) and in transmitting the information through long-lived self-configuring embedded networked sensing systems. In brief, these networks

OCR for page 187
190 Appendix B embed a computational intelligence in the environment, linking sensor pods through wireless technology in a manner that allows the network to conduct adaptive monitoring and real-time control. Development of new sensors from nanosensors to new satellite-based systems was also described in Chapter 2. (7) Merging observations with models: Data sets are frequently assimilated into models, both to provide model-based forecasts (e.g., upper air observations used for weather forecasting; precipitation and river stage observations to fore- cast flood stages) and to predict variables not well measured (e.g., nonpoint pol- lution runoff, terrestrial evaporation). (8) Using results from an integrated observational-modeling system: Data and model products have no value unless they are used. They can only be used if they can be easily discovered, acquired, and understood in a timely manner to those who wish to apply them to practical issues such as flood forecasting, water availability modeling, and ecological flows, as inputs to decisionmaking. The communication and delivery of data and information to such end users is the back-bone to a beneficial integrated system. New “web-based hydrologic ser- vices” are being developed at the University of Texas by Professor David Maidment under National Science Foundation funding, and similar applications with remote sensing at the University of Illinois by Professor Praveen Kumar (Box 3-1). These nascent activities facilitate data discovery, acquisition, and integration need to be further developed and integrated with users through dem- onstration projects; and other competing approaches need to be developed and evaluated through similar means.