Letter of Intent to Propose - University of Colorado Boulder



A. Abstract

Transforming climate information into usable knowledge

to enhance decision-making in water resources management of the Río Salado del Sur Basin, Argentina

NOTE: This proposal is being simultaneously submitted by the University of Miami (Dr. Guillermo Podestá, PI). Although the text describes the entire set of tasks proposed by both universities, the budget figures given here only include the University of Colorado component.

Lead Investigator at Boulder, CO : Rajagopalan Balaji,

Dept. of Civil Env. And Arch. Engineering, University of Colorado, Boulder, CO, USA

Telephone: 1.303.492.5968; FAX 1.303.492.7317, E-mail: balajir@colorado.edu

Total project cost: $ 402,416 - Total University of Colorado budget: $ 200,437

Project performance period: 1 May 2007 – 31 October 2009 (2.5 years)

We propose to explore conditions for the effective use of weather and climate information (from weather and seasonal climate forecasts to plausible projections of decadal trends) to address issues related to management of flood risks, ground water recharge, and water supply in several urban concentrations in the Río Salado del Sur Basin in central-eastern Argentina. The Río Salado watershed is subject to marked climate variability on both interannual and decadal scales that has resulted in frequent flooding events affecting many urban regions in the basin. Floods also have affected associated transportation infrastructure and the agricultural production (about 25% of total crop and cattle production in Argentina) that supports the economies of affected urban regions. Most urban concentrations in the basin depend on ground water to supply their population therefore it is important to understand the interactions between surface water and ground water availability.

We submit that, in order to be useful, climate information must be relevant to viable decisions and compatible with current decision processes. Further, the use of climate information must be understood within a broader framework that includes, as an example, the economic, social, cultural, and institutional contexts. Finally, early collaboration with stakeholders and organizations with operational mandates for the management of water is crucial to ensure the effective transition from science-derived knowledge to action. Therefore, the work we propose involves two major streams:

i) Development of tools to “translate” weather and climate predictions into distributions of relevant outcomes and useful information products with appropriate temporal and spatial resolution; and

ii) Facilitation of the process of transference of tools and findings to governmental institutions in the target region with operational responsibilities for management of water resources and risks from extreme climatic events.

Stakeholder agencies will be fully involved in both streams of the proposal to ensure from the outset that the tools envisioned and the framework developed are relevant and compatible with the goals and objectives of water resources management. Federal and state-level agencies in Argentina such as the Undersecretariat for Water Resources and the Dirección Provincial de Saneamiento y Obras Hidráulicas already have indicated their willingness to collaborate with us in this project, and specific letters of commitment are included in the proposal. We emphasize that many of the tools proposed already have been developed with funding from prior and ongoing NOAA and NSF projects, and have been in use for agriculture management. In this proposal we intend to synthesize these existing tools into a framework that is amenable to water resources management.

B. Results from Prior NOAA Support

Building capacity to use climate information and forecasts to enhance decision-making in agriculture: An application to the Argentine Pampas. G. Podestá (PI), K. Broad and H. Herzer, NOAA Environment, Science and Development Program, May 2004.

This project aims to enhance the match of informational climate messages to the characteristics and decision contexts of Argentine farmers and their technical advisors. The project involves two components. First, we assessed the missions, functions, capabilities, and products of two boundary organizations (institutions with missions involving both producers and users of climate information): the Argentine Meteorological Service and AACREA, a non-profit farmers’ group. In a second component, we have conducted 60 interviews with farmers in two regions of Argentina (one climatically optimal, the other more marginal) to assess what farmers already know and believe about climate, and what knowledge they lack. A few results can be highlighted:

• Argentina’s National Meteorological Service lacks functions required by a user-centric boundary organization. For instance, no formal mechanisms exist to receive user feedback. Provision of climate services takes a backseat to the production of daily weather forecasts. As a result of our findings, in May 2006 the Met Service implemented a Web-based survey (that project participants helped design) about its climatic products. About 250 responses have been received and will be analyzed shortly.

• A promising approach involves strategic partnerships between the Met Service and other organizations (governmental or private) with a clear boundary role in climate-sensitive sectors. For example, we are facilitating a process of collaboration between the Met Service and a non-profit group of Argentine farmers (AACREA) whose mission is technology dissemination. As a consequence of our intervention, a dialog has started to implement the dissemination through AACREA of weather information relevant to farmers produced by the Met Service.

• We worked with Met Service staff in the operational implementation for the Pampas of the Standardized Precipitation Index (SPI), an index developed to monitor the onset and duration of droughts in the American West. SPI maps can be found at . With our assistance, Met Service staff wrote an article describing the SPI that will appear in the July 2006 AAACREA Magazine (distributed to about 1700 farmers in Argentina), and developed a tutorial on the index to be posted in the WWW in July 2006. These efforts clearly demonstrate our commitment to foster the transition towards operational implementation and dissemination of NOAA-sponsored research results.

• Our “mental model” interviews helped to identify misconceptions and erroneous beliefs that need to be addressed in subsequent communication efforts. For example, a location A may have better crop yields than location B because of better soils or better agronomic management. Nevertheless, farmers often assume that “climate in location A is more favorable than in B”, assigning to climate any advantages in other environmental or technological factors. That is, the causes of good agricultural results frequently are confounded.

• Another interesting example of misconceptions about climate involves a reverse assignation of causality: if a given season is unusually dry (wet), farmers assume that a La Niña (an El Niño) event is happening, even if these extreme El Niño Southern Oscillation (ENSO) phases, associated with rainfall anomalies in spring/summer, are not really occurring. This reverse assignation of causality may affect the perceived salience of truly ENSO-related climate anomalies.

Publications: Two manuscripts are in preparation describing each project component. A poster describing the Standardized Precipitation Index was presented in the Latin American Congress of Meteorology (Buenos Aires, October 2005). An abstract using the SPI to describe recent drought events in the Pampas will be presented at the Argentine Agrometeorology Congress (September 2006).

Decision-making in agricultural production in the Argentine Pampas: Alternative choice process formulations and the value of climate information. E. Weber (Columbia Univ.), D. Letson (Univ. of Miami) and G. Podestá. NOAA Human Dimensions of Climate Change Research Program. October 2004.

Specific objectives of this project include the empirical identification of objective functions of a sample of Pampas’ farmers (as well as their technical advisors) and an assessment of the prevalence of decision objectives outside the subjective expected utility model that is conventional in economics. We have implemented four alternative objective functions: (a) subjective expected utility, (b) prospect theory’s value function, and (c-d) utility and values corrected according to regret theory. We have used crop simulation models to find optimal land allocations for each function. We are currently computing estimates of the economic value of climate information for expected utility and prospect theory’s value function. Ultimately, understanding individual goals will allow us to tailor agronomic technical advice to a farmer’s personality characteristics and economic context. Some preliminary results can be highlighted:

• Simulations show differences in the actions identified as optimal by each tested objective function.

• We have explored the parameter space for each of the implemented functions, and found regimes where results (optimal actions) change quickly in response to small variations in parameter values. These results will help plan subsequent field studies.

• A discontinuity in the original formulation of the prospect theory value function makes it difficult to optimize this function using widely available software packages (e.g., GAMS). We have developed an alternative formulation that does not have the discontinuity problem and is otherwise equivalent to the original.

Publications: Three papers have been submitted on (a) the implementation of an alternative formulation for prospect theory’s value function, (b) a unified parameterization of corrections to utility by regret and disappointment, and (c) simulation of optimal actions for various psychologically plausible objective functions. Several presentations about this work have been made in different fora.

Understanding the Spatio-Temporal Variability of the North American Monsoon: Implication to Water Resources Management in the South Western US. R. Balaji (PI), E. Zagona, M. Clark, S. Gangopadhyay (Co-PIs), NOAA Office of Global Programs.

This is an ongoing project to (a) understand the space-time variability of North American summer monsoon; (b) develop forecast tools and; (c) demonstrate the utility of incorporating the variability information into water resources management. So far we found the following interesting aspects: (i) There seems to be a significant shift in the timing of the monsoon– i.e., monsoon occurring later in recent decades; (ii) consequently, the July-August rainfall is now pushed into September, which shows a strong increasing rainfall trend; (iii) Land conditions in the pre-monsoon (late winter, early spring) seems to be a key player in the monsoon timing and strength; (iv) The pre-monsoon land conditions are strongly modulated by ENSO, thus preceding winter ENSO might have a substantial influence on the following summer monsoon, than previously thought and; (v)This raises hopes for long-lead forecast of the summer monsoon that can be of great importance to resources management.

Publications. We anticipate 3~4 peer-reviewed papers, the first of which, describing the space-time variability of the North American monsoon is in press in Journal of Climate special issue on North American Monsoon. In addition, we have presented these results at the Spring 2005 AGU in a special session on North American Monsoon organized by the PI.

C. Statement of Work

1. Introduction

1.1 Statement of the Problem

Current use of weather and climate information to support decision-making in sensitive sectors of society is evolving rapidly. Improved scientific and technological capabilities, a growing appreciation for the importance of climate on human endeavors (including sustainable development and poverty mitigation), and an increasing demand for information to support decisions are all providing strong incentives for the production and delivery of useful climate and weather data, information, and knowledge to decision-makers (National Research Council, 2001; Dutton, 2002). The ability to provide timely information offers an exciting opportunity to learn how important and prevalent weather- and climate-sensitive sectors such as water resources management may respond.

Several empirical studies have identified theoretical and practical obstacles to the use of weather and climate information (Mjelde, 1998; Pulwarty and Redmond, 1997; Finan, 1998; Stern and Easterling, 1999; Roncoli et al., 2001; Broad and Agrawala, 2000; Broad et al., 2002; Lemos et al., 2002; Patt and Gwata, 2002; Ziervogel, 2004; Rayner et al., 2005). The obstacles are diverse, ranging from limitations inherent to the climate system’s complexities (for example, seasonal forecasts have coarse spatial and temporal resolution, not all relevant variables can be predicted, the skill of forecasts is not well characterized or understood, contradictory predictions may coexist), to procedural, institutional, and cognitive difficulties in receiving or understanding the information, or in the ability and willingness of decision-makers to modify their actions. In the case of water resources, additional impediments include the fragmented nature of institutional responsibilities, fairly rigid operational procedures, and an environment generally averse to technical innovation or experimentation (Rayner et al., 2005).

We argue that at the root of most barriers or impediments to the use of weather and climate information lies a fundamental misfit between the capabilities and communication abilities of producers of the information, and the expectations, needs and beliefs of potential users of such information. To overcome this misfit, the informational message must be matched to the characteristics and situation of the target group (Stern and Easterling, 1999). For example, Orlove and Tosteson (1999) stress that climate forecasts must be well matched to the problem frame, decision-making processes, and capacity for adaptive response of the users. Nevertheless, the fit between information and needs is not simply the combination of a predetermined set of users’ potential decisions and the climate information. Rather, the fit develops from the interaction over time of information producers and users, in which users learn to expand their choice set in response to the availability of new information, and producers adapt their information products to the evolving capacity of users (Patt, 2000; Pulwarty and Melis, 2001).

1.2 Brief Project Overview

We propose to implement and transfer to operational users a suite of tools to “translate” raw weather and climate information into relevant products or indices of decision variables to enhance the management of risks linked to flooding and water supply in cities within the Río Salado del Sur Basin (hereafter, Salado Basin) in central Argentina. The Salado del Sur is a component of the Río de la Plata Basin that drains a large portion of South America. The Salado Basin is located in the flat plains of central Argentina known as the Pampas, one of the major agricultural areas of the world, and is the site of considerable beef and crop production. The basin encompasses almost half of the Province (state) of Buenos Aires, the most densely populated and economically important province of Argentina. Precipitation and streamflow in the Salado watershed show marked interannual and decadal climate (see “Background” section).

Important climate-related water management issues in the Salado Basin are associated with urban flood prevention and mitigation, and with water supply to urban populations. Further, the economies of most cities in the basin are tightly linked to agricultural production in surrounding areas (many farm owners reside in nearby cities but rely on farm-derived income; businesses in small urban areas provide most services and supplies to local agriculture). For this reason, droughts have impacts on the prevalently rainfed agricultural systems of the region that, in turn, influence considerably the economic welfare of adjacent urban regions.

To address potential mismatches between the capabilities of weather and climate information producers and the needs and expectations of users, we plan a two-pronged approach. As part of the first prong, we propose to adapt and combine existing tools and approaches (e.g., downscaling procedures, hydrologic models) to “translate” raw weather and climate information into a suite of relevant products or indices of decision variables useful to support adaptive management responses to climatic risk factors. The products initially envisioned will allow stakeholders to address issues such as “Is there flooding risk to city X as a consequence of predicted weather 3-5 days ahead?”, or “What are the chances that the supply of drinking water to city Y may face shortages as a result of expected dry conditions during next season?”, or even “Will city Z require deeper wells if water table depths increase as a consequence of decadal changes in climate?” We stress that the envisioned list of products and indices (see “Proposed Approach”) undoubtedly will be revised after early interactions with local collaborators and stakeholders.

There is abundant evidence that the effective use of weather and climate information is not simply the transmission of salient technical facts, but also involves a process of understanding and melding political, social, institutional, and other contextual interests that often may conflict (Pulwarty and Melis, 2001; Wernstedt and Hersh, 2002; Lemos et al., 2002; Varis et al., 2004). For this reason, the second prong of the project will involve close collaboration with, and outreach to a range of stakeholders (from federal and state governmental water resource organizations to operators of urban water systems) to identify and address issues that could impede or facilitate use of weather and climate information, and to foster a smooth transition towards operational use of the planned products. Several aspects of the work proposed here may be highlighted:

• The importance of weather and climate variability on the management of urban water resources and flooding risks will be simultaneously explored for a broad range of temporal scales, from synoptic weather to seasonal-to-interannual and decadal scales.

• The project will build on pre-existing tools, some of which (e.g., watershed models) already are in use by relevant stakeholders. Nevertheless, we will generate a relevant suite of products and indices unavailable so far and that will contribute to management of climate risks in urban regions.

• Investigators in the project are fully committed to, and have a proven track record (see “Results from Prior NOAA Support”) of fostering the transition of research results towards operational use. We plan to work closely with governmental, non-governmental and private stakeholder institutions at international, federal, provincial, and city levels (see “Supplementary Documentation”) to transfer and help implement the tools and procedures developed during the project.

2. Background

2.1 The Study Area

The Río Salado del Sur Basin encompasses about 170,000 km2 in the Province of Buenos Aires, the most densely populated and economically important province of Argentina. The river originates at El Chañar Lake, near the border with the Province of Santa Fe, and flows southeastward for about 650 km before reaching the Atlantic Ocean at the Bay of Samborombón, about 170 km south of the city of Buenos Aires (capital of Argentina). Average streamflow near the ocean is about 88 m3 s-1. Figure 1 shows the Province of Buenos Aires and the boundaries of the Salado Basin. The basin has three main sub-regions: the Salado-Vallimanca region, the Northeast region, and the western “chained lagoons” region. The latter two sub-regions became part of the basin only recently, after construction of the Oeste and Alsina drainage canals. The basin population of about 1.3 million is almost totally dependent upon agricultural production for its livelihood, with most of the area devoted to cattle ranching but with an increasing amount of cropping in some parts. There are about 25 smaller cities (10,000-50,000 inhabitants) and five cities with more than 50,000 people (Junín, Chivilcoy, Olavarría, Azul and Tandil; Figure 1).

The highest elevations in the basin (500-1000 m) are found in the southern part of the basin (Tandil and Sierra de la Ventana). Most of the basin, however, is a smooth plain with little relief. Thus, drainage systems in the region are poorly developed and the proportion of precipitation that ends as river runoff is only about 10% (López et al., 2003). The low-energy Salado system does not have much capacity to respond to high precipitations. As a result, the basin suffers periodic flooding, with significant impacts on urban and road infrastructure, and agricultural production (which has considerable influence on urban economies). Flooding has been more frequent since the 1970s, when annual precipitation in the Pampas increased.

As a result of concerns about recurrent floods in the Salado Basin, the Government of Argentina commissioned a study by the British consulting firm Sir William Halcrow & Partners. The resulting report identified 14 urban areas in the Salado Basin that had high vulnerability to floods: San Miguel del Monte, Chascomús, Azul, Rauch, Ayacucho, Olavarría, Tapalqué, Mones Cazón, Navarro, Dolores, Pehuajó, Carlos Tejedor, Carhué-Guaminí, and Roque Pérez. The Province of Buenos Aires recently has initiated a process of revision and update of the 1997 Halcrow document and has involved various Argentine universities and research institutions in the development of a Master Plan to decrease the vulnerability of urban regions and agricultural production in the Salado Basin. The process explicitly addresses the inclusion of climate information in the development of the Master Plan. For this reason, we will join efforts with partners in Argentina to explore how weather and climate information can be considered to protect urban concentrations from flooding and ensure sustainable supply of water to their populations.

2.2 Climate Variability in the Study Area

El Niño Southern Oscillation (ENSO) is the major single source of climate variability on seasonal-to-interannual scales in southeastern South America (Ropelewski and Halpert, 1987, 1989; Grimm et al., 2000; Montecinos et al., 2000; Boulanger et al., 2005). In the target region, during November-December, El Niño events are associated with higher median precipitation and likelihood of high rainfall anomalies than other phases, whereas La Niña events show markedly lower median rainfall and a narrower range of anomalies (Podestá et al., 1999; Vargas et al., 1999; Rusticucci and Vargas, 2002). Predictions of rainfall over southeastern South America during October–December exhibit one of the highest skill levels anywhere (Goddard et al. 2002).

Several studies have addressed the interannual variability of streamflows in the Plata Basin (e.g. Aceituno, 1988; Mechoso and Perez-Iribarren, 1992; Marengo, 1995; García and Vargas, 1998; Genta et al., 1998; Robertson and Mechoso, 1998; Bischoff et al., 2000; Camilloni and Barros, 2000; Robertson et al., 2001; Berbery and Barros, 2002; Berri et al., 2002; Krepper et al., 2003). In addition to the interannual signal, the region’s climate shows marked decadal-scale variability. A steady increase in annual precipitation has been observed since the 1950s over most of central-eastern Argentina (Krepper at al. 1989; Castañeda and Barros 1994; Rusticucci and Penalba 2000). Streamflows for the Paraná, Paraguay, Uruguay, and Negro rivers show a decrease during the first half of the 20th century followed by an increase in the second half, most marked since about 1970 (García and Vargas, 1998; Genta et al., 1998; Robertson and Mechoso, 1998). Robertson et al. (2001) analyzed the interannual to decadal predictability of the Paraná River extracting near-cycling components of the summer river streamflow. They found that the ENSO oscillatory component was associated to changes in the probability distribution of monthly flows and that the decadal modulation of ENSO may be important. More details on the variability of the Río de la Plata basin climate and hydrology can be found in a recent report by the VAMOS Scientific Study Group (2001) and the references therein.

3. Proposed Approach

3.1 Overview

The overall goal of this project is to develop a framework for the effective use of weather and climate information to enhance decision-making and address resource management and decision-making in the Río Salado del Sur Basin. Three main design considerations will guide the project: (a) to draw as much as possible on existing insights, methods and tools, (b) to integrate seamlessly approaches from the natural and social sciences, and (c) to work closely with appropriate stakeholders to foster operational implementation of the tools and products developed. The work proposed involves two major prongs.

The first prong of the project will focus on the production and assessment of usable weather and climate information and products (sensu Lemos and Moorehead, 2004). The rationale for this component is that information on sector-relevant variables or indices (e.g., flood risk maps) often is more important or relevant to stakeholders than climate information. As part of this component, weather and seasonal climate information will be linked with downscaling tools and hydrological process models. The hydrological model will receive climatic variables such as precipitation and temperature, and will output estimates of surface and ground water flows, including estimates of water table depth (relevant to assess risks to water supply for urban regions in the Salado basin, which rely mostly on ground water). The result will include a suite of relevant prognostic hydroclimate information and products derived from weather and climate forecasts and from plausible projections of decadal climate trends.

The second prong of the project will center primarily on the “demand side” of weather and climate information. Our goals are to develop better matches among what is needed, what is asked for, what is provided, and what actions can be taken (Pulwarty and Melis, 2001). Introducing weather and climate information into real-world, well-accepted decision procedures may be an effective strategy (Rayner et al., 2005). For this reason, we will work with a range of stakeholders, from federal and state governmental water agencies in charge of flood control to operators of city water systems, to characterize decision processes and ensure that weather and climate information can be built-in into existing procedures. A highlight of this project is that rules for the operation of flood control structures under construction (e.g., at the Mar Chiquita Lagoon, intended to decrease flood risks to the city of Junín) have not been defined yet. Therefore, we have an unusual opportunity to work with agencies tasked with operating these structures to ensure that climate information is included from the outset into operational procedures. In this component, we also will explore possible institutional, legal and infrastructural constraints to use of weather and climate information. Finally, we will strive to develop an understanding of the decision contexts within which trade-offs in water-related decisions take place.

Subsequent sections describe in detail the two main components of the proposed work and its sub-components. The underlying project framework is presented in Figure 2.

3.2 Specific Tasks Proposed

3.2.1 Data Compilation

Historical Climate Data: Collaborators at the Argentine Meteorological Service (SMN, after its acronym in Spanish) will compile quality-controlled historical climate series (maximum and minimum daily temperature, daily precipitation) for several locations in or around the Río Salado Basin and the period 1950-present. These data will be used to “train” stochastic weather generators, calibrate and validate hydrologic models, and compute indices that require long records.

Weather Forecasts: We will use weather forecasts 3-5 days into the future routinely produced by SMN, including fields of numerical model output used to develop the forecasts. These forecasts will be combined with tools to produce ensembles of equally likely weather series that, in turn, will serve as input to watershed models in order to produce, for example, maps of flood risks.

Seasonal Climate Forecasts: Seasonal climate forecasts from operational agencies can range from the prediction of an El Niño or a La Niña event (that in turn influence regional climate conditions) to probabilistic statements about the likelihood of regional precipitation or temperature falling within certain categories (e.g., “above normal”, “below normal”). Temperature and precipitation forecasts for the study region will be obtained for the International research Institute for Climate Prediction (IRI), regional agencies (e.g., Brazil’s CPTEC) and Regional Climate Outlook Fora, periodically held in the region (Buizer et al., 2000).

Plausible Decadal Climate Scenarios: The marked increase since the 1970s in spring-summer precipitation in central-eastern Argentina has contributed to significant changes in land use patterns and frequency of floods in the Salado Basin (Castañeda & Barros 1994; Viglizzo et al 1995, 1997; Satorre 2001). Although, at present, predictions of decadal variability have very little skill, we will define various plausible scenarios (we stress these are not predictions) to explore interesting situations.

3.2.2 Translating prognostic climate information into relevant hydrological information

Prognostic weather and climate information in the Salado Basin is available on two main temporal scales. On one hand, weather forecasts are available for short time scales (say, up to a week into the future). On the other hand, agencies such as CPC, IRI, or ECMWF routinely issue large-scale climate forecasts on a seasonal or monthly time-scale. Each of these types of prognostic information may assist in different types of decisions, but both sources require a process of “translation” into usable scenarios before they can be used in conjunction with process models (e.g., a watershed model) to produce relevant hydrological products.

Seasonal forecasts typically are probabilistic and categorical. Rainfall is often forecast as the chance of being ‘above normal’, ‘below normal’ or ‘near normal’. Each of these categories corresponds to a tercile of the distribution (terciles include 1/3 of the sorted historical distribution of precipitation totals for a given period). For example, a hypothetical forecast for Oct-Dec precipitation in the Salado Basin may assign probabilities of 0.45, 0.35, and 0.20 to precipitation totals in the lower, middle, and upper terciles, respectively. Because the probability of rainfall in the lower tercile is 0.45 (as opposed to the climatological 0.33), this forecast implies that dry conditions are more likely than normal or wet conditions, although the latter two also may occur. This probabilistic nature needs to be emphasized so that users do not assume it to be a deterministic prediction that is wrong when the actual climate is not as forecast (Ziervogel and Calder, 2003).

Seasonal forecasts typically are issued for areas that have sizes of the order of hundreds of square kilometers (see Buizer et al., 2000, for examples of forecasts). Similarly, output from the regional Eta model, while improving upon the coarse spatial resolution of the GCMs, still provides results on a grid larger than the point-scale required to drive watershed models. The key issue for both types of forecasts, therefore, is how coarse forecast information can be translated into weather scenarios at observation points in the basin and, subsequently, into hydrologic (i.e., streamflow attributes) scenarios that can be used in water resources management.

3.2.2.1 Generating synthetic daily weather sequences consistent with seasonal climate scenarios

This section describes how we plan to generate ensembles of synthetic daily weather scenarios conditioned on a large-scale seasonal climate forecast using stochastic weather generators. These scenarios drive the watershed model to produce ensembles of streamflows and other decision variables. Weather scenarios are generated using stochastic weather generators and these are used routinely in water, agricultural, and erosion control management (Skidmore and Tatarko, 1990; Wilks, 1997; Dubrovsky et al., 2000). Most approaches are based on parametric or semi-parametric schemes (i.e., underlying distributions are assumed and described via parameters estimated from historical data). Recent developments of stochastic nonparametric weather generators and their successful applications (Rajagopalan and Lall, 1999; Yates et al., 2003; Clark et al., 2003) are an attractive alternative to traditional parametric approaches. The nonparametric methods avoid subjective judgments about appropriate model forms and probability distributions (rarely tested formally). Furthermore, as data-driven methods, they can capture deviations from theoretical probability distributions and nonlinearities in the associations between variables (Wilks and Wilby, 1999; Rajagopalan and Lall, 1999). The number of parameters to be estimated increases significantly with the number of weather variables and desired periods (weeks, months, etc.); with short historical records, this can lead to unstable models. In contrast, non-parametric methods are more parsimonious.

The nonparametric weather generator based on the K-nearest neighbor (K-NN) framework (Lall and Sharma, 1996; Rajagopalan and Lall 1999; Yates et al. 2003) has been shown to be simple and effective at simulating the distributional properties of the historical weather, also, it can be easily extended to multisite weather generation (see e.g., Yates et al., 2003). In this approach, all weather variables are considered simultaneously as members of a feature vector for day t. The algorithm selects the next day t+1 starting from day t from a set of potential nearest neighbors in the historical record. The simulated weather sequences will reproduce all the statistical properties (i.e. probability density function, PDF) of the historical data. This unconditional weather generator has been applied and tested at several locations in the USA (Yates et al., 2003; Rajagopalan and Lall, 1999) and also in the Pampas region in Argentina by the PIs.

Recently (Apipattanavis, 2006) the PIs modifed the K-NN framework to a two-step semiparametric approach, where in, a Markov Chain was used to generate the precipitation state (i.e. wet or day) and the K-NN approach described above to generate the weather sequences. This has been shown to perform much better in capturing the wet and dry spell statistics, which is key in streamflow simulation. They demonstrated the utility of this approach in the Pampas region in Argentina. Furthermore, Apipattanavis (2006) also developed a modification of this two-step semiparametric approach to generate weather sequences at multiple sites, conditioned on categorical seasonal climate forecasts such as those issued by IRI. They successfully applied this in the Pampas region and demonstrated its application in crop yield simulation and forecast. This framework was developed as part of an ongoing NSF project. In this proposal we will use this framework to generating daily weather sequences at multiple sites for use in streamflow generation.

The PIs have extensive experience with development of the weather generator and also its application to several locations and sectors. Given this, the testing and validation of the stochastic weather generator in the Salado Basin will be done in short order.

3.2.2.2 Generating synthetic daily weather sequences consistent with short-term weather forecasts

The methodology described in the preceding section can be adapted to produce weather scenarios at observation points in the basin which can, subsequently, be converted into hydrologic (i.e., streamflow) attributes. The K-NN weather generator described above can be easily modified to translate operational weather forecasts into weather scenarios conditioned on the forecast.

The required modification of the K-NN algorithm was developed and successfully demonstrated on two climatologically different basins in the US by Gangopadhyay et al. (2005). In this approach, they downscaled the NCEP (1998) medium range forecast (MRF) model output to point-scale weather scenarios. The algorithm queries days similar to a given feature vector (i.e., the vector of output from regional numerical prediction models) in the archive, and using EOF (Empirical Orthogonal Functions) analysis identifies a subset of days (K) similar to the feature day. These K days are then weighted using a bi-square weight function, and randomly sampled to generate ensembles. The feature vector could be precipitation output at the grids or circulation variables (winds, pressures etc.), because the models are generally good at capturing the circulation fields compared to precipitation. Gangopadhyay et al. (2005) show very good skill from this downscaling approach. The PI (R. Balaji) was involved in developing this approach. We will apply this approach to the Salado Basin.

3.2.2.3 Generating synthetic daily weather sequences consistent with plausible decadal climate scenarios

As mentioned earlier, there has been an increase in precipitation in the Salado Basin since the 1970s and hence changes in land use patterns and increased flooding. For long term watershed management and development we need to generate plausible decadal climate scenarios and consequently, using the watershed model corresponding flow scenarios for use in planning. To this end, K-NN stochastic weather generator can be easily modified. Blocks (say 30-years long) of consecutive historical years are selected at random and the weather generator applied to the selected block to generate ensembles of precipitation and temperature. These will then drive the watershed model to obtain streamflow scenarios in the basin.

3.2.2.4 Translating synthetic daily weather into hydrological scenarios

The stochastic weather sequences consistent with weather/climate forecasts and decadal scenarios will drive a watershed model system to generate ensembles of streamflow and other hydrologic conditions in the basin, thus providing their probability distributions. As part of the Salado Master Plan, our collaborators at the School of Engineering of the University of Buenos Aires (FIUBA) are collaborating with the state-level Dirección de Saneamiento y Obras Hidráulicas (DIPSOH) to develop models of the Salado Basin. Preliminary but promising modeling results are illustrated in Figure 3 that shows modeled and observed Salado River streamflow and water table depth near Junín.

The preliminary results shown in Figure 3 were produced by a model developed using the MIKE-SHE modeling system developed by the Danish Hydraulic Institute (DHI, ). MIKE-SHE is a physically-based, distributed-parameter integrated tool that can simulate the entire land phase of the hydrologic cycle, from rainfall to runoff, evapotranspiration by vegetation, infiltration and ground water flows. The model allows the simulation of hydraulic structures. MIKE-SHE already is operationally used by DIPSOH, therefore the use of this environment increases the probability of adoption of the tools we develop. Although MIKE-SHE is a commercial product with an average cost of about 10,000 Euros (about 13,000 US dollars), the commercial representative of DHI Water and Environment in Argentina has agreed to donate a free license to the University of Buenos Aires for use in this project if NOAA funding is awarded (see “Supporting Documentation”). This represents significant leveraging of NOAA funds.

The integrated watershed model system will be completely implemented and calibrated during the first year of this project. Output from the watershed model system will include ensembles of streamflow at several locations, water depths, potential flooding, soil moisture in the basin etc. All of these variables form a rich array of decision inputs that can be used by decision makers (described in the following section).

3.2.2.5 Using hydrological scenarios to generate relevant hydrological products and indices

The hydrologic model will be driven by the daily weather sequences generated from the stochastic weather generator described earlier. Ensembles of output from the hydrologic model will be obtained at both seasonal (based on seasonal climate forecasts) and 1-2 week time scales (based on short term weather forecasts). The ensemble outputs at the seasonal time scale will help in resource planning (such as water supply management, conservation measures etc.) at the start of the season and the outputs at the short term will enable emergency planning (such as flood evacuation). To this end, two types of products will be obtained from the proposed integrated framework.

(i) Outputs from the hydrologic model.

The model provides the following key variables spatially throughout the basin.

• Streamflow;

• Soil moisture;

• Depth of the water table;

(ii) Derived or value-added products.

From the model outputs listed above, we will develop value-added products that can be directly mapped to decision making. Two such outputs are:

• Flood risk, including flood persistence;

• Risk of water supply shortages.

Stakeholder’s local knowledge and input and knowledge will be critical in developing the value-added products, including the definition of important risk thresholds. The risk estimates, obtained spatially over the entire basin, will be overlaid on a GIS (coverages are under development as part of the Salado Master Plan), thus providing an effective visualization of the decision-making information.

Flood risk will be expressed as inundation maps for different recurrences (1 time every N years), which will be obtained by constructing the envelope of maximum flow height during a flood wave passage for a series of stations along the stream (Re et al., 2004). Flood persistence, in turn, will be calculated for each station as the time interval during which the flow height stays above a predefined threshold; the flood persistence map will be built as contour lines for the flood persistence. For urban areas, flood risk maps will be useful as a base to establish land use regulations and, eventually, define structural protection needs. This information should be used by municipalities as part of urban planning and land zoning, and by the Provincial government when defining big defense works.

Maps of the depth of the water table at regional scale will allow an estimation of the buffering capacity of the unsaturated zone for the precipitation scenarios to be considered in the project. Operators of water supply systems in urban regions can use this information to assess risk of shortages in water supply, or the cost of providing water (pumping ground water from a deeper water table is costlier).

3.2.3 Ensuring Information Relevance and Fostering Operational Use

The second major component of the project will center on the “demand side” of weather and climate information to ensure better matches among what is needed, what is asked for, what is provided, and what actions can be taken (Pulwarty and Melis, 2001). As the section title indicates, the two major goals of this component are (a) to ensure that the products and indices to be produced are relevant and consistent with users’ needs and decision processes (b) to facilitate the transition of research results to operational use. To achieve both goals, we will partner with appropriate stakeholders with operational missions related to the regulation and operation of water in the Salado Basin. The stakeholders include federal and state water agencies in charge of flood prevention and mitigation and operators of water systems in the basin’s cities. The latter include a range of organizations from public agencies to private concessions to cooperatives of city/town inhabitants.

3.2.3.1 Mapping the decision-making landscape

Adaptive responses to weather and climate information require that a range of options exist (Podestá et al., 2002). Therefore, as a first step towards assessing the scope for adaptation, we will construct decision maps and calendars in close collaboration with stakeholders. These maps will characterize (a) decisions related to flood prevention and mitigation, and urban water supply, (b) the timing of such decisions, and (c) a range of realistic options and constraints for each decision. This effort has three simultaneous goals. First, understanding stakeholders’ decision processes will help ensure that information is produced that is relevant and compatible with users’ needs and decision procedures. Second, we aim to identify “entry points” for climate information (Pulwarty and Melis, 2001). Third, we want to assess the degree of awareness or concern about weather/climate risks among stakeholders, as previous studies have found that feeling at risk from these factors stimulates the decision to incorporate related information to the decision process (O’Connor et al., 2005).

We will use non-probabilistic sampling of stakeholders and snowball interviewing (i.e., asking initially identified stakeholders to suggest other interviewees). We will use influence diagrams (Morgan et al., 2002) to build a conceptual model of climate-related risks and other context factors, as well as the multiple objectives influencing flood control and mitigation and water supply decisions. Influence diagrams show associations between variables or factors (e.g., “A influences B”) and provide a very effective participatory mechanism to develop a consensus description of the target system, as most people intuitively understand “bubbles and arrows” diagrams. Simple models of decision processes and the contextual factors influencing them have advantages for addressing complex systems initially, as they make assumptions transparent and uncertainty more traceable (often not the case for larger, unwieldy models). Simple initial descriptions can evolve into increasingly complex models.

As a way of triangulating the appropriateness of the decision maps/calendars and the conceptual model of the context of decision-making, we envision a series of role-playing exercises in which stakeholders will be asked to make specific decisions in the light of hypothetical information. For example, operators of urban water systems may be given various scenarios of expected seasonal conditions, and asked to make realistic decisions (e.g., what kind of actions to take to prevent water shortages). Public administrators may be asked to explore the consequences of various flood scenarios on urban and road infrastructure. In both cases, process models (watershed models) will provide stakeholders with immediate feedback on the results of their decisions, facilitating adaptive learning.

3.2.3.2 Towards operational adoption of research-derived knowledge

A major goal of the project is to foster a smooth transference of research results to operational agencies. To achieve this goal, we will rely on lessons learned during previous NOAA-sponsored research in Argentina. The major lesson we learned is that stakeholders’ trust is essential. To earn their trust, we will elicit frank input and feedback from key stakeholders (e.g., technical personnel in provincial water agencies) from the outset of the project. Another lesson is to understand stakeholders’ constraints, motivations and incentive structure. In plain language, we must understand how much time/effort different stakeholders can commit to interactions with scientists, and ensure that they get involved when it is really crucial. Also, we must provide tangible incentives such as co-authorship in scientific papers or leadership in development of highly visible products (as we did with the Met Service and precipitation indices). Funds are being requested to support an MS/PhD student at the University of Colorado to work on the implementation of the proposed tools. We hope to find a motivated, talented candidate among the Argentine collaborators or stakeholder organizations to provide yet another channel of communication with the US scientists.

Another lesson we have learned in previous NOAA work is that an appropriate mechanism to facilitate use of climate and weather information is to encourage strategic partnerships between producers and users of the information. In this case, we will facilitate dialogues between staff at the Argentine Met Service and the various water organizations involved in this project. We hope that this effort will achieve two major goals. On one hand, stakeholders will establish direct contact with people who operationally produce and disseminate weather and climate info. They will get a chance to learn first-hand what the capabilities and limitations of the current state of the art in forecasting. This is important because in Argentina there are multiple sources of climate forecasts (issued by university groups, individual scientists) and the pedigree of these forecasts is not always clear. On the other hand, Met Service personnel do not often meet users of their products, and our experience has been that they get highly motivated when they learn that their information (in particular, non-traditional products) is put to practical use.

Users who are educated in the meaning and significance of weather and climate information will probably make greater and better use of this information. For this reason, special emphasis will be placed on the transfer of knowledge to administrators and technicians in governmental agencies such as DIPSOH, and to water supply operators. Although we plan to maintain an active, collaborative working relation with many key stakeholders, there is a need for an active outreach component targeted on those stakeholders not actively involved in the research. We plan to provide periodic briefings to and elicit feedback from the less-actively involved stakeholders. A project WWW site will be used to disseminate findings and provide a forum for long-distance research collaboration.

3.2.3.3 Towards a decision support system

To facilitate progress towards formalized procedures to incorporate weather and climate information into water management in the Salado Basin, we will start incubating the development of a Decision Support System (DSS). To this effect, we have asked Dr. E. Zagona to join this project. Dr. Zagona is an expert on Water Resources Decision Support System and also the director of the Center for Decision Support for Water and Environmental Systems (CADSWES, ), sponsored by the USBR. CADSWES has developed DSS tools for water management that are widely used by water managers, USBR, municipalities and a variety of stakeholders. Furthermore, these tools have enabled stakeholder participation in decision making in the Western US – which is a significant step in the participatory decision making. CADSWES also has agreed to host the graduate student (hopefully from Argentina) that will be working on this project, which will enable close collaboration and development of lasting professional ties.

We feel that developing a DSS on the Río Salado del Sur Basin would be the culmination of the process of incorporating climate information and stakeholder participation. As mentioned earlier, some hydraulic structures to regulate flows in the Salado are being constructed. Operational rules for these structures are not in place yet, and we hope to contribute to their definition, including the consideration of weather and climate information. Tools such as those developed by CADSWES would facilitate the participatory development of operational rules.

3.2.3 Monitoring Project Progress and Operational Adoption

As stated above, this project will develop and implement a series of hydroclimate products that so far have not been widely available (if at all!) to the user community. This offers unique opportunities not only to design the products in a collaborative mode to ensure relevance and usefulness, but also to monitor the process of diffusion of these new products. Agrawala and Broad (2002) suggest different models of technology transfer as a tool for situating the state of seasonal forecast dissemination and adoption. Ziervogel and Downing (2004) have studied networks of stakeholders involved in the diffusion of seasonal climate forecasts.

We will work with stakeholders to monitor periodically awareness and use of the various products to be implemented as part of the project. If necessary, short surveys will be administered once a year to stakeholder organizations. On the last year of the project, we will engage a local social scientist (Dr. Cecicilia Hidalgo, University of Buenos Aires) who is collaborating with us in an NSF project in which she is leading the assessment of interdisciplinary collaboration and interactions among researchers. Dr. Hidalgo will be asked to conduct an independent assessment of the stakeholders’ experience of collaboration and the adoption of project results.

4. Synergies with Ongoing Efforts

The tools to be used in this research, particularly, the weather generators and downscaling techniques were developed with NOAA-OGP funding (especially, the Western Water Assessment, WWA). Outcomes from this project will provide useful experience to the water management efforts in the Western US. The project also will benefit significantly from interactions with a NOAA-sponsored project that, although focused on agriculture in Argentina, is exploring existing mental models of climate variability in the study region, as well as existing institutional structures for production and dissemination of climate information. Further, investigators in this team are participating in NSF-sponsored projects exploring adaptive management in complex natural/human systems and decision-making under climate uncertainty. Results from the multiple perspectives offered by these complementary projects will provide rich insights.

5. Project Management

This proposal is being simultaneously submitted by the University of Colorado (Dr. Balaji Rajagopalan, PI) and the University of Miami (Guillermo Podestá, PI). Drs. Rajagopalan and Podestá will share overall responsibility for project completion. Dr. Angel Menéndez (PhD in Hydraulics, Iowa State University) is a Professor at the School of Engineering of the University of Buenos Aires and also is affiliated with Argentina’s National Water Institute. He is involved in research on the Salado basin and is a consultant to national and international (e.g., the Plata Basin Inter-Governmental Committee) water organizations. He will participate actively in this project and will supervise a post-doctoral researcher in Argentina who will be dedicated full time to this project. Table 1 shows the major groups of tasks and the investigators who will lead each component.

Table 1. Project work plan, indicating major groups of tasks, responsibilities, and timing of each task.

|RESEARCH COMPONENTS |Year 1 |Year 2 |Year 3 |

|Data Preparation and watershed model: testing/calibration of MIKE-SHE watershed model | |

|(Menendez/Rajagopalan); data compilation (entire team) | |

|[pic] |[pic] |

G. Investigators’ Vitae

Biographical Sketch

RAJAGOPALAN BALAJI

A. PROFESSIONAL PREPARATION

Kurukshetra University, India, B. Tech. (with honors) in Civil Engineering, 1989

Indian Statistical, Calcutta, India, M. Tech. (with honors) in Operations Research and Quality Reliability, 1991

Utah State University, Logan, UT, USA, PhD in Civil Engineering (Stochastic Hydroclimatology), 1995

B. APPOINTMENTS

2000 (Aug) - present Assistant Professor, Department of Civil, Environmental and Architectural Engineering, University of Colorado at Boulder.

2001 (June) – present Fellow, Co-operative Institute for Research in Environmental Sciences (CIRES, University of Colorado, Boulder, CO.

2000 (Aug) – present Adjunct Associate Research Scientist, International Research Institute (IRI), Lamont-Doherty Earth Observatory (LDEO), Columbia University, NY

1999 (July) – 2000 (August) Associate Research Scientist, International Research Institute (IRI), Lamont-Doherty Earth Observatory (LDEO), Columbia University, NY

1997 (July) – 1999 (June) Associate Research Scientist, Lamont-Doherty Earth Observatory (LDEO), Columbia University, NY

1995 (April) – 1997(June) Post-Doctoral Research Scientist, Lamont-Doherty Earth Observatory (LDEO), Columbia University, NY

1991 (Oct) – 1995 (April) Graduate Research Assistant, Utah Water Research Laboratory, Utah State University, Logan, UT

C. HONORS AND AWARDS

Distinguished Utah State University Dissertation in Engineering: 1993-1995. Utah State University nomination for the Council of Graduate Schools Distinguished Dissertation Award.

Honorable mention, 1996 Award for the Outstanding Water Resources Dissertation in the field of Engineering and Physical Sciences,The Universities Council on Water Resources.

Nominated for the 1996 Lorenz G. Straub award for the most meritorious thesis in hydraulics and hydrology and finished in the top three.

Young Researcher award: 2003. Department of Civil Environmental and Architectural, University of Colorado, Boulder, CO.

Participated in the 15th Annual Beckman Frontiers of Science Symposium, National Academy of Sciences, Irvine, CA, Nov 6 – 8, 2003.

Research Development award: 2006. Department of Civil Environmental and Architectural, University of Colorado, Boulder, CO.

D. RESEARCH INTERESTS

Stochastic Hydrology and Hydroclimatology; Nonparametric functional estimation techniques (probability density Functions, regression, scenarios generation, forecasting); Understanding low frequency climate variability and its signatures on regional hydrology; Incorporating climate information in water resources/hydrologic decision making; Understanding spatio-temporal variability in Indian summer monsoon; Stochastic modeling of hurricane tracks; Nonlinear Dynamics - recovering dynamics from data; Bayesian techniques for optimal combination of information from multiple sources and decision making.

E. PUBLICATIONS

(i) Publications most closely related to the proposed project

KrishnaKumar, K., B. Rajagopalan, and M.A. Cane, On the weakening relationship between the monsoon and ENSO, Science, 284, 2156-2159, 1999.

K. Krishna Kumar, M. Hoerling and B. Rajagopalan, Advancing Indian Monsoon Rainfall Prediction, Geophysical Research Letters,  32, L08704, 1-4, 2005.

Yates, D., S. Gangopadhyay, B. Rajagopalan, and K. Strzepek, A nearest neighbor bootstrap technique for generating regional climate scenarios for integrated assessments, Water Resources Research, 39(7), 1199, 2003.

Gangopadhyay,  S., M. Clark, B. Rajagopalan, Statistical Downscaling Using K-Nearest Nieghbors ,Water Resources Research, 41, W02024, 1-23, 2005.

Singhrattna, N., B. Rajagopalan, M. Clark and K. Krishna Kumar, Forecasting Thailand Summer Monsoon Rainfall, International Journal of Climatology, 25, 649-664, 2005.

(ii) Other significant publications

Rajagopalan, B., U. Lall, and M. A. Cane, Anomalous ENSO occurrences: an alternate view, Journal of Climate, 10(9), 2351-2357,1997.

Rajagopalan, B., U. Lall, and S. Zebiak, Optimal Categorical Climate Forecasts through Multiple GCM Ensemble Combination and Regularization, Monthly Weather Review, 130, 1792 – 1811, 2002

Rajagopalan, B., E. Cook, U. Lall, B. Ray, Temporal Variability of ENSO-drought association in the South West US, Journal of Climate, 13, 4244-4255, 2000.

Grantz, K., B. Rajagopalan, M. Clark and E. Zagona, A Technique for incorporating large-scale climate information in basin-scale ensemble streamflow forecasts  Water Resources Research, 41, W10410, 1-13, 2005.

Singhrattna, N., B. Rajagopalan, K. Krishna Kumar and M. Clark, Interannual and Interdecadal Variability of Thailand Summer Monsoon, Journal of Climate, 18, 1697-1708, 2005.

F. SYNERGISTIC ACTIVITIES

Associate Editor, Geophysical Research Letters and ASCE - Journal of Hydrologic Engineering

Member, The American Geophysical Union

Member, Precipitation Committee, AGU Hydrology section.

Reviewer, Water Resources Research, Science, Geophysical Research Letters, Journal of Hydrologic Engineering, Journal of Climate, Advances in Water Resources, Tellus, NSF, NOAA and NASA proposals.

Organized a session titled Low frequency climate variability signatures on regional hy drometeorological variables - implications to hydrologic forecasting and planning at the Spring meeting of AGU, Boston, May, 1998.

Organized a session titled Incorporating climate variability information in water resources decision making”, at the Fall AGU, San Fransisco, Dec 2002.

G. COLLABORATORS & OTHER AFFILIATIONS

(i) Collaborators

Clark, M., CIRES/Univ. of Colorado at Boulder; Kushnir, Y., Columbia Univ. at NYC; K. Krishna Kumar, Indian Institute of Tropical Meteorology, Pune, India; Lall U., Columbia Univ. at NYC; Hoerling, M, Pulwarty, R., CDC-NOAA/Univ. of Colorado at Boulder; Katz, R., NCAR, Boulder; Podesta, G., Univ. of Miami, Miami; Zagona, E., CADWES/Univ. of Colorado at Boulder; Zebiak, S., IRI/Columbia Univ. at NYC;

(ii) Graduate and Postdoctoral Advisors

Upmanu Lall, Columbia University, NY, Mark Cane, Columbia University, NY.

(iii) Theses Advised (at University of Colorado, Boulder)

MS David Newmann; James Prairie; Katrina Grantz; Nkrintra Singhrattna; AdamHobson and J. D. Emmert

PhD Yeonsang Hwang; Somkiat Apipattanavis; Paul Block

(iv)Advisors for Research Associates

Dr. Subhrendu Gangopadhyay, University of Colorado, Boulder, CO, 2002 -

Dr. Krishna Kumar, Visiting Fellow, CIRES/University of Colorado, Boulder, CO 2003-2004

Biographical Sketch

EDITH A. ZAGONA, Ph.D., P.E.

Director

Center for Advanced Decision Support for Water and Environmental Systems (CADSWES)

Department of Civil, Environmental and Architectural Engineering

Campus Box 421, University of Colorado at Boulder, Boulder, Colorado 80309-0421

(303) 492-2189, E-mail:zagona@cadswes.colorado.edu

Academic Degrees

Ph.D., Civil Engineering, University of Colorado, 1992 Thesis: “Model-Predictive Control of Automated

Canals”

M.S.C.E., Hydraulics, Colorado State University, 1983

B.S., Civil Engineering, University of Arizona,1978

B.A., Philosophy and Mathematics, University of Arizona, 1975

Appointments

Since March 2001 Interim Director, CADSWES.

Since September 1992: Research Associate, CADSWES. Principal Investigator of Research and development of RiverWareTM since 1993.

January 1988 - August 1992: Professional Research Assistant, CADSWES. PI of research grants with Imperial Irrigation District associated with doctoral research.

February 1986 - March 1987: Design and Planning Coordinator for the Central Arizona Project, U.S. Bureau of Reclamation, Arizona Projects Office, Phoenix, Arizona, 1985-87.

July 1978 - February 1984: Hydraulic Engineer, Water Conveyance Branch, U.S. Bureau of Reclamation Engineering and Research Center, Denver, Colorado, 1978-1984.

Recent Research and Development Activities

Principal Investigator for research and development of RiverWareTM, a river basin simulation and optimization model licensed internationally by the University Technology Corporation. Since 1993, grants totaling approximately $12 million with Tennessee Valley Authority, Electric Power Research Institute, U.S. Bureau of Reclamation and U.S. Army Corps of Engineers. Lead team of 15 to 20 professional researchers, water resources engineers, software developers and student assistants.

Recent Publications

Neumann, D., B. Rajagopalan, and E. Zagona, " A regression model for daily maximum stream temperature" (accepted) ASCE Journal of Environmental Engineering, July 2003.

Wheeler, K., T.M. Magee, T. Fulp, and E. Zagona, " Alternative Policies on the Colorado River," Proceedings of University of Colorado Natural Resources Law Center Conference: Allocating and Managing Water for a Sustainable Future: Lessons From Around the World, Boulder, Colorado, June 2002.

Zagona, E.A., Fulp, T.J., Shane, R., Magee, T. and Goranflo, H.M. “RiverWare: A Generalized tool For Complex River Basin Modeling,” Journal of American Water Resources Association, August 2001.

Magee, T, Zagona, E. and Frevert D., "Operational Policy Expression and Analysis in the RiverWare Modeling Tool," to appear in Environmental and Water Resources Institute’s (EWRI’s) World Water & Environmental Resource Congress, ASCE New York, NY, 2001.

Eschenbach, E., Magee, T., Zagona, E., Goranflo, M. and Shane, R. “Multiobjective Daily Operations of Reservoir Systems via Goal Programming,” ASCE Journal of Water Resources Planning and Management, March 2001.

Zagona, E. and Magee, T., “Modeling Hydropower in RiverWare,” in “Waterpower ‘99, Proceedings of the International Conference on Hydropower, ASCE New York, NY, 1999.

NATIONAL ENVIRONMENTAL POLICY ACT STATEMENT

The research activities that will be carried out in this project is analysis and synthesis of observed climate and hydrological data. These datasets are readily available to the PIs. This work will be carried out on the Rosenstiel School campus of the University of Miami, at the University of Colorado at Boulder, and in collaborating institutions in Argentina. The work will make no impact on species or habitat. It includes no construction activities. There are no environmental concerns.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download