Linking Activities, Expenditures and Energy Use into an Integrated Systems Model to Understand and Predict Energy Futures

Subhro Guhathakurta in collaboration with Eric Williams and Eric Hittinger at Rochester Institute of Technology

Funded by NSF grant # 2243100

Predicting and managing energy demand are crucial tasks for addressing climate change and other environmental impacts of energy use. The mainstream models of energy demand are reductionist, dividing demand into separate categories such as residential, commercial, and transportation, and analyzing each separately. This research will develop a holistic model of energy demand, one that considers how consumer actions affect multiple sectors at the same time, including residences, vehicles, commercial buildings, server/data networks, and the production of purchased goods. The model includes consumer ownership and use of energy- consuming technologies such as vehicles or home furnaces, and accounts for how people use time and spend money. 

The new model will be constructed for the U.S. by integrating government micro-data on consumer behavior (American Time Use Survey, Residential Energy Consumption Survey, National Household Travel Survey, Consumer Expenditure Survey), using modern data analysis methods to integrate them. The integrated dataset will provide information about energy device ownership and use, internet use, time spent in commercial buildings, and expenditures on goods. A set of models will map these consumer attributes to energy use and carbon emissions. The holistic model will help understanding of the broader effects of demand interventions. For example, how are the carbon benefits of electric vehicles affected by induced changes in consumer purchases and activity choices? The model will help assess the effect of behavioral changes not typically considered in policy (e.g. encouraging telework), and thus could broaden the scope of policy options considered.

Heat Risk Surveillance

Brian Stone

The UCL Heat Tolerance Index (HTI) tracks the incidence of combined levels of outdoor temperature and humidity at which an individual will experience a rise in core body temperatures from exposures of 1-2 hours. Recent physiological studies have found the threshold temperature for adverse health impacts – measured as the wet bulb temperature (a combination of heat and humidity) – to be lower than previously established. For a healthy young adult, a wet bulb temperature of 87°F, equivalent to 87°F with 100% relative humidity (or 100°F with 60% relative humidity), is found to induce a fever within 60-90 minutes of exposure at a moderate activity level. With sustained exposures (> 3 hours), core body temperatures may exceed 104°F, elevating the risk of heat stroke. We refer to any days with a temperature in excess of this 87°F limit as threshold heat events. The HTI quantifies the difference in maximum annual wet bulb temperatures and the 87°F threshold at which outdoor activities cannot be safely continued. A second index estimates the remaining time to threshold exceedance based on a continuation the five-year average rate of increase in the annual maximum wet bulb temperature. 

NSF S&CC: Design and Development of a Near Real-Time Community Crowdsourced Resilience Information System for Enhancing Community Resilience in the Face of Flooding and other Extreme Events

Subhro Guhathakurta in collaboration with Peng Chen (CSE) and other researchers at the University of South Florida

Funded by NSF Smart and Connected Communities Award # 2325631

Communities along the coast are increasingly vulnerable to coastal hazards such as flooding due to extreme weather events and sea level rise. In the US alone, 40% of the population lives in coastal cities and subjected to elevated risks of such hazards. The probability of a flooding event in these communities is also increasing with global warming. The proposed project will design a neighborhood-scale Community Resilience Information System (CRIS-HAZARD) by leveraging citizen science and community participation for enabling real-time data-driven decision-making to make communities more resilient to flooding. CRIS-HAZARD will support frequent bi-directional flow of information among communities, research scientists, and decision-makers. The objective is to develop a platform that facilitates the smart and connected city framework by engaging diverse communities to improve the lives of all citizens, especially those who are marginalized. The project is piloted in Pinellas County, Florida, in the Tampa Bay region on the Gulf Coast of west-central Florida. This region’s geography and low elevation make it especially vulnerable to climate change-induced extreme weather events like flooding.

Unlike previous attempts at integrating data and models to predict flooding events, the approach of CRIS-HAZARD is distinctive as this research pioneers the integration of user-supplied data (crowd-sourced and social media) with real-time flood prediction models and uncertainty analysis techniques, which is expected to advance our understanding of risk and resilience in coastal communities facing persistent flooding events. The initiative integrates the expertise of research institutions, government agencies (Office of Emergency Management or OEM), local stakeholders, and community engagement networks to enhance community-based planning and policy decisions, promoting community resilience. Furthermore, the project fosters customized resiliency planning at the neighborhood level engaging citizen scientists as partners. It aligns with the National Science Foundation's mission to provide transparent and accessible information on risks and vulnerability, contributing to the development of smart and resilient communities nationwide.

A pilot experimental project for predicting pedestrian flows using computer vision and deep learning

Subhro Guhathakurta in collaboration with Uijeong Hwang at the Atlanta Regional Commission

Funded by US Department of Transportation Center for Understanding Future Travel Behavior and Demand at UT Austin

Walking for transportation, health, and pleasure is an essential part of people’s lives in most cities. Knowing where people linger, the destinations that attract them, and how those places are accessed could assist in optimizing business locations and providing better security. In addition, predicting and sharing congestion times and locations (perhaps in real-time as in Waze for cars) could also provide useful information to travelers who can then choose appropriate travel routes and improve travel efficiency. Yet, we know far less about the spatial and temporal variations in pedestrian volumes than we know about vehicular movement.

Goals:

• Estimate pedestrian flows from video footage of pedestrians at different locations.

• Determine how different layers of information (count, direction, etc.) improve the pedestrian flow models.

Ensuring Fair and Equitable Funding of Rural Transit in Georgia after the 2020 Census

Subhro Guhathakurta (Co-PI) in collaboration with Prof. Laurie Garrow (PI)

Funded by the Georgia Department of Transportation1

Rural public transit systems are typically small, demand-responsive systems. Revenues are generally not sufficient to cover the system’s costs, and the FTA § 5311 and § 5304 programs provide capital, planning, and operating assistance in support of these systems. State DOTs are responsible for developing and executing a process to fairly and equitably distribute federal transit funds to rural public transit systems.  In FY21, more than $728 million in federal funding was allocated nationwide for rural transit and GDOT distributed $25 million to 85 rural transit operators. These operators provide demand-responsive transit service for 114 (out of 159) counties.

The objective of this research is to calculate 5311 and 5304 funding appropriations at a county level for FY23-FY30. We will use GIS-based tools, data from Census – including the new rural and urban designations that will be released in December of 2022 – and the National Transit Database to calculate annual federal rural transit funding for each county in Georgia.  Our calculations will use FTA’s appropriation flowcharts for the 5311 and 5304 programs and FTA’s data value tables.

The Role of On-Demand Transit in a Just Future Public Mobility (JustFuture)

Subhro Guhathakurta in collaboration with Prof. Vanessa S

Funded by Vinnova – the Swedish Innovation Agency

The objective of JustFuture is to examine, compare, and learn from two On-Demand Transit solutions deployed in different urban and rural contexts in two continents; MARTA Reach in Atlanta and Plusresa in the region of Skåne in southern Sweden. The aim is to explore the role of on-demand transit in issues of transport equity and equality. The project goal is to lay a strong foundation for a long-term sustainable and productive collaboration between involved project partners in the US and Sweden.

Atlanta Climate Vulnerability Map

Brian Stone in collaboration with Evan Mullen

Funded by City of Atlanta

The Urban Climate Lab worked with the City of Atlanta to create a climate vulnerability map for all neighborhoods in the city — the first of its kind for a US city. Using the interactive map, residents can search for their neighborhood to see how it ranks in comparison to the other 247 neighborhoods in the city. Residents can also zoom down to the level of individual city blocks to see how their location compares to others in the city. Map colors correspond to the average daily high temperature in July (°F). Click on the button in the top right corner of the map to access the temperature range for each color class.

Climate risk statistics can be accessed by clicking on any neighborhood. All neighborhoods are ranked by heat risk, flood risk, and overall climate vulnerability (with lower numbers indicating higher risk). Additional statistics include the typical daily high temperature in July, the percentage of neighborhood homes with air conditioning, the neighborhood tree canopy coverage percentage, the number of trees needed to reach 50% tree canopy coverage (a goal adopted by the City Council), and how neighborhood heat risk would change if a goal of 50% tree canopy coverage was achieved.

Neighborhood statistics include the relative heat-related mortality of the neighborhood compared to the median mortality rate for the city. This metric is calculated as the percentage difference between the median mortality rate (deaths per 100,000 residents) for the city and that of the neighborhood of interest. Reported neighborhood statistics also include the heat island intensity experienced in the neighborhood. This metric is calculated as the difference in the maximum July temperature in the neighborhood minus the maximum July temperature in the coolest neighborhood in the city. For example, a heat island intensity of 6 would indicate that the average maximum temperature for the neighborhood is 6°F warmer than that of the coolest neighborhood. An overview of this climate risk resource can be found in the Atlanta-Journal Constitution.