The study area is the Connecticut River watershed, extending from the U.S.-Canada border north to the Long Island Sound south. The watershed drains over 28,490 Km2 and contains 390 towns and cities with a population of approximately 2.3 million. The watershed has 79% forest cover and 11% agriculture (Marshall and Randhir, 2008b). Modeling the dynamics of FEW systems requires using multiple spatial and temporal processes of watershed hydrology, built energy, and agriculture subsystems. A coupled framework based on patterns and processes in FEW systems helps understand the integration among modules, linkages between multiple scales and temporal processes, identifying mitigation opportunities, and studying interactions between agriculture, built environment, and water resources. The conceptual model of the proposed analysis presents the flow of information in the integrated assessment. Integrating various models will involve synthesis into a tradeoff analysis, subbasin statistics of FEW subcomponents, and a multi-attribute optimization model to be integrated into a decision framework. The research will combine spatial analysis (GIS), Energy models (EnergyPlus and DesignBuilder), agricultural simulations (APEX and IFSM), GCM outputs, SWAT, and LULC models at multiple scales (building, raster (site), land use/soil combination (virtual basin), urban region, subbasin, and watershed) and long-term time scales (100 years – an ensemble of GCMs). Climate and LULC change can be linked to smaller time scales (daily, seasonal, and yearly - SWAT) (Marshall and Randhir, 2008b), thus becoming relevant for FEW systems research. The SWAT model will be used in this study to predict water resource impacts under baseline climate and land use. The calibrated model will be subject to ensemble climate scenarios and land-use changes as drivers to study changes in runoff, groundwater recharge, sediment loading, nutrient loading, and water yield. Data and model inputs will be developed using the SWAT model interface from the EPA’s Better Assessment Science Integrating Point and Nonpoint Sources (BASINS) model (USEPA, 2015). The energy used in the built environment of the watershed will be modeled using built environment energy simulations with EnergyPlus (USDOE, 2016) and DesignBuilder (DesignBuilder Software Ltd, 2016). For agricultural simulations, we will use the APEX model (Williams et al., 2012) for the watershed. Specific dairy farms will be simulated using IFSM (Rotz et al., 2015). The MCA (Markovian, Monte Carlo Analysis) will be used to develop a matrix of probabilities for each land class transition. For climate change assessment, this study will use the most recent evaluation from an ensemble of GCM models compiled by the IPCC-AR5– Fifth Assessment. We will use qualitative and quantitative methods to study attitudes toward FEW systems. Semi-structured interviews and workshops (interactive GIS/modeling) will be used to assess the nature of attitude toward green infrastructure. A survey instrument will quantify attitudes toward green spaces and information on socioeconomic, demographic, geographic, and local characteristics. Dillman's method (Dillman, 2000) will be used to develop high-quality questionnaires and higher response rates and will be administered by both mailed and Internet-based questionnaires. The specific format and content will be developed through consultation with the Advisory Panel that will be formed during the start of the project. The survey will be cleared through IRB at the University of Massachusetts before contacting respondents. A focus group will be used to pretest the questionnaire for fine-tuning the structure and finalizing the questions for the survey. All survey participants will be sought full consent and notified of their right to withdraw. No private information will be collected that could identify the individual or group. The FEW system integrations will be accomplished through subbasin assessment of individual outcomes through spatial integration of attributes to derive tradeoff information.
Tim
Randhir
