2022 BTAA GIS Conference

The 2022 BTAA GIS Conference has concluded.

The 2022 Conference was held on November 15th and featured presentations, lightning talks, a map gallery, and career panels.

Conference Recordings

View the 2022 conference recordings here or on YouTube.

Agenda

9:00 - 9:10 CST (10:00 - 10:10 EST) Welcome

material-clock-time-nine-outline: 9:10 - 9:50 CST (10:10 - 10:50 EST) Lightning Talks

Lightning Talks

1. Range Shifts Predicted with Climate Change in a Common North American Songbird

Susan Reed, Indiana University Bloomington

With a net loss of 3 billion North American birds since 1970, and 2.5 billion being native migratory birds, it is critical to investigate population dynamics and threats to species that impact abundance 1 and geographic distribution across space and time. Predicting species range in response to variation in climate is a goal shared among the fields of both conservation and evolutionary ecology. Species distribution models (SDMs) are visual representations of simulated spatial and temporal patterns of a species in response to predictor variables (PVs), making them useful tools for predicting a species’ range for extrapolated future climate scenarios. As rapid environmental change is experienced, animals must adapt to changes in their environment at a minimum of the same rate, lest they risk extinction. Range shifts and adjustments in movement strategies are mechanism of coping with climate change, and SDMs give insight into how these mechanisms may play out in natural populations—information that may be used to forecast population dynamics and persistence. A better understanding of where, when, and how species distributions are shaped provides valuable information to conservation practitioners, evolutionary ecologists, and policy makers for decision-making regarding habitat and species management. Here, I use ArcGIS Pro2 software to demonstrate SDMs of climate niches for a common migratory North American songbird, the slate-colored dark-eyed junco (Junco hyemalis), utilizing contemporary and extrapolated climate scenarios to model changes in its range as it experiences climate change.

2. Estimating Urban Rooftop Solar Power Production Potential In Chicago

Damian Hamielec, Purdue University

Anecdotally, Chicago has a lot of underutilized rooftop space. Consider that Chicago is extremely dense with thousands of buildings, it also has that many roofs that are rarely used for patios, gardens/green roofs, or solar installations. This analysis looks to answer how much energy could Chicago generate from rooftop PV installations that are placed in the most optimal locations throughout the city limits.

3. Developing A Predictive Cultural Heritage Model of Michigan

Amanda Tickner, Michigan State University

We present a predictive model for cultural heritage management and preservation that incorporates historical features as a basis for the development of a predictive model for decision making. This presentation will describe our development process and future possibilities for expansion of our model.

4. Deep learning-based tree species identification using UAV RGB images*

Yunmei Huang, Purdue University

Information at the species level is necessary for precision forest management and biodiversity conservation. However, it remains a challenging task with traditional remote sensing techniques because of the limit in imaging resolution. Unmanned aerial vehicles (UAVs), as an emerging RS technique, are getting more attentions for their low-cost and high-resolution advantages. However, most of the studies focusing on UAVs images for tree species identification are based on multispectral, hyperspectral images or combined with LiDAR data. The gap exists in using UAVs RGB data to identify tree species at the stand-level in forests with various species, especially for temperate forests. In this study, we built a species dataset using high-resolution canopy images (3-5 centimeters) collected by drones from different seasons. Transfer learning techniques were used to test the performance on specie identification using ResNet18, YOLOv5, and EffcientNetb0, which achieved accuracies above 98% among 4 species including northern red oak, black cherry, and white oak. Testing how flight height affects species classification using RGB images from different flight altitudes is one of the projects we are working on. Tree species identification among more species and individual tree detection from natural forests will be part of our future work.

5. How would SES and neighborhood environment influence individuals' physical inactivity and psychological distress in the City of Chicago?*

Bing Han, Purdue University

I propose to explore how adult physical inactivity and serious psychological distress are influenced by individual socio-economic status and neighborhood environment by using GIS data and survey data on the Chicago Data Portal. Previous research has repeatedly found that family socio-economic status is related to individuals’ health through multiple pathways, like nutrition intake, social and emotional functioning, and adoption of health risk behaviors. Meanwhile, community environments, like community crimes, green spaces, and access to public health services, are also important to people’s health. In medical sociology, the interacting effects of social factors on the individual level and resources on the community level on individuals’ health have been explored. However, it usually lacks a view of geography due to the limitation of data or methodology skills. This study aims to apply GIS to the field of medical sociology to better understand the pattern of how health is influenced. The City of Chicago provides resourceful data on health and neighborhood characteristics. I propose to combine the Healthy Chicago Survey on individuals living in poverty and the environmental data regarding green spaces, crimes, race/ethnicity segregation, education, etc. to explore is there is a moderating, say a protective, effect on community/neighborhood level environment on the negative effect of low socioeconomic on individuals’ health outcomes, including physical inactivity and serious psychological distress.

Session 1: Justice and Social Geography

10:00 - 11:00 CST (11:00 - 12:00 EST)

Reconstructing Historical Neighborhoods from historic Sanborn Maps Using Machine Learning

Harvey Miller, Center for Urban and Regional Analysis / The Ohio State University; Yue Lin, CURA/Ohio State; Ningchuan Xiao, CURA/Ohio State; Gerika Logan, CURA/Ohio State

The building of the interstate highway system in US cities in the 20th century split and sometimes destroyed entire neighborhoods, mostly those housing African Americans, immigrants and other minorities. Planners of the system routed some highways, often purposely, through vulnerable and deprived neighborhoods occupied by people of color. Once thriving and full neighborhoods, these neighborhoods were split and, in some cases, even fully demolished for these urban highways. This caused an economic downturn in many of these neighborhoods that lasts until the present day. Remaining residents still live with the negative consequences of these highways, including disconnections, poor air quality, noise and road trauma. The Ghost Neighborhoods of Columbus project will highlight these lost and damaged neighborhoods by using machine learning/artificial intelligence techniques to extract building level data from historic Sanborn fire insurance maps to populate a GIS database, allowing 3D visualizations of the lost urban landscapes, and support analysis of the wealth, communities and activities lost to urban highway construction. This presentation will discuss our workflow and show results from Driving Park and Hanford Village in Columbus, Ohio – two neighborhoods damaged by the construction of interstate highway I-70.

Using maps to communicate the relationship between broadband and health

Jess Hoffelder, University of Wisconsin Population Health Institute; Anne Roubal, University of Wisconsin Population Health Institute

Broadband was highlighted as a critically important service during the pandemic. Although it is widely documented that internet availability is inconsistent across rural areas, pockets of neighborhoods in urban areas also lack access. We used the example of Milwaukee, Wisconsin a historically segregated city to show how Federally sanctioned residential segregation in the 1930’s is still reflected in health inequities in socioeconomic health metrics today. We used ArcGIS Pro to map several community socioeconomic metrics such as children in poverty, homeownership rates, and those who have completed high school, as well as historically redlined neighborhoods, against access to high-speed broadband to show the spatial patterns. We then transformed these maps into several GIFs, shared in an article on the County Health Rankings and Roadmaps website and social media platforms, to help users better overlay the different metrics. We find, unsurprisingly, that census tracts with lower percentages of broadband access are correlated with those in which more children live in poverty, fewer individuals own their homes or graduated high school, and were more likely to be redlined in the past. These findings highlight the existence of digital redlining and point to the need to reduce the inequity in broadband access to avoid further harm to communities experiencing disinvestment.

Overcoming the challenges of developing a geospatial database for longitudinal study of urban childhood obesity

David Tulloch, Rutgers University; Kristen Lloyd, Rutgers University; Michael Yedidia, Rutgers University; Derek Delia, Rutgers University; Robin DeWeese, Arizona State University; Francesco Acciai, Arizona State University; Punam Ohri-Vachaspati, Arizona State University

Childhood obesity is a complex problem to which a child's environment contributes significantly. Translating the dynamic urban realities of both the food environment and physical activity (PA) environment into consistently measurable variables is a challenging GIS problem, especially in cities with limited municipal data infrastructure. Using proximity as an indicator of access, we created an unique longitudinal dataset for this project which led to a variety of important findings about childhood obesity and access to parks and corner stores, perceptions about safety impacting walkability, and the value of developing an early habit of walking or biking to school. Since 2008, our research team built a routinely updated database representing the dynamic food environment of Camden, Newark, New Brunswick and Trenton, New Jersey. Updates included mapping corner stores, groceries, fruit and vegetable store, bakeries, butchers, supermarkets, as well as fast-food and full-service restaurants. While general food outlet data is acquired from various sources, to be valid each year’s entry needs to be checked with direct contact and using supplementary information sources (e.g. external locational data). Simultaneously, the team mapped the physical activity environment, including parks and PA facilities (like gyms). The database also includes upgrades to community PA features (like parks, trails), extracting information from city agencies, project grants and other public sources. This presentation will provide a brief overview of the key contributions of GIS to this 15-year collaboration of multidisciplinary multi-institutional childhood obesity research. It will then focus on the geospatial challenges encountered and lessons learned.

Session 2: Climate and Carbon

11:15 - 12:00 CST (12:15 - 1:00 EST)

Facilitator: Karen Majewicz, University of Minnesota

Detecting forest carbon storage by remote sensing

Bowen Li, Purdue University; Guofan Shao, Purdue University; Michael Saunders, Purdue University

Forest is a major contributor in carbon sequestration and reduces impacts from climate change. Carbon storage quantification is useful in evaluating and adjusting forest management activities. Although traditional field measurement such as destructive sampling can accurately estimate carbon on individual trees, it’s very time-consuming and labor-intensive. Remote sensing method, however, uses aerial photos to capture forest change within a short period of time. The extraction and integration of spatial (e.g. canopy height) and spectral information (e.g. greenness) can be used for biomass estimation which eventually converts to carbon.

Modeling Present and Future Ecosystem Services and Environmental Justice within an Urban-Coastal Watershed

Marci Meixler, Rutgers University, New Brunswick, NJ; Max R. Piana, USDA Forest Service, Northern Research Station, Amherst, MA; Alexis Henry, King County Department of Natural Resources and Parks, Seattle, WA

Globally, urban-coastal areas are expected to experience substantial landscape shifts as a result of climate change induced sea level rise. Such changes will impact valuable ecosystem services. We developed a model in the urban-coastal Jamaica Bay, New York watershed to quantify present-day carbon sequestration, aboveground carbon storage, and belowground carbon storage ecosystem services and predict the impact of sea level rise through 2100 on future ecosystem services. Due to the disproportional impact of sea level rise on wetlands and natural coastal-fringe habitat, we predicted that carbon sequestration and storage would be significantly impacted in these habitats. Our model projected that future carbon sequestration, aboveground carbon storage, and belowground carbon storage potential in our watershed will have losses up to 0.16%, 15%, and 51%, respectively. We paired our present-day ecosystem service model results with socio-economic data and identified clusters in which mitigation action may help to address impacts of sea level rise and environmental justice. Our work addresses the need for better understanding of urban- coastal ecosystem service flows and the potential impact of future landscape change on environmental and socio-economic factors. Our results provide support for actions to increase urban-coastal resilience through informed design, planning, and management of these ecologically and socially significant landscapes.

12:00 - 1:00 CST / 1:00 - 2:00 EST - LUNCH BREAK

Keynote - Our Home on Native Land: Connecting People and Stories using Indigenous Geographies

1:00 - 2:00 CST / 2:00 - 3:00 EST

Facilitator: Joshua Sadvari, Ohio State University

Tanya Ruka, Native Land Digital

Tanya Te Miringa Te Rorarangi Ruka is a Māori Indigenous artist and designer living in Te Whānganui-a-Tara, Aotearoa (Wellington, Capital City of New Zealand). The land of 3 tribal nations: Te Atiawa, Ngati Toa Rangatira and Taranaki Whānui ki te Upoko o te Ika.

She is of Ngati Pakau, Te Uriroroi, Te Parawhau, Te Mahurehure Ngapuhi, and Waitaha descent. She is a member of the Taheke Gazetted Māori Committee of the National Māori Council. She has a Master of Art and Design, and works as an independent indigenous researcher, she is active in environmental issues from an indigenous perspective in Aotearoa and globally. Working with the Waitaha Executive Grandmothers Council she has been documenting and mapping the sacred stories of her Tupuna ancestors as evidence for tribal environmental issues involving land and water protection claims through the Treaty of Waitangi. As a Māori artist she is inspired by her ancestry, and the creation stories that place the land as ancestor, and the Master Navigator voyagers of the Pacific who always kept their eyes on the horizon in hopeful anticipation, bringing the tribe safely to land. Tanya works with Mātauranga Māori (ancestral knowledge and navigational tools) to design pathways of transitional Indigenous Futures and Indigenous Speculative Design. Working in the digital realm creating stories that firmly place indigenous concepts, knowledge, perspectives and ways of being in the landless territories of new imagined futures. She is currently working with dedicated indigenous and non-indigenous textile researchers, academics, scientists, engineers, growers and local Iwi (tribes). Documenting the journey to develop circular designed, native plant fibre materials and textiles that will help to connect people back to the land through indigenous ways of knowing. She is developing a Community Rongoā (Māori native plant medicine) Forest initiative weaving together ancestral knowledge with integrated app based technology.

Tanya hopes to continue the legacy of her late Uncle, who was sent by her Grandmother into the world to open up pathways of communication with other indigenous tribal nations. As Research Communications Lead for Nativeland Digital she is honoured to be a part of the team and is dedicated to the representation of indigenous tribal voices and their homelands.

Rudo Kemper, Native Land Digital

Rudo is a non-native geographer and technologist working in solidarity with Indigenous and other local communities to co-create and use digital tools for self-determination and self-representation. He got his start working with communities in the Amazon rainforest in the early 2010s, where he accompanied communities like the Matawai Maroons of Suriname in mapping and monitoring their ancestral lands, and documenting their traditional knowledge and oral histories. Rudo currently works with Digital Democracy, where he is accompanying communities across the globe in building and using mapping tools to defend their lands, and stewarding the development of the Earth Defenders Toolkit, a collaborative digital platform for earth defender communities and their allies. Originally from Curaçao, Rudo has worked with communities in Suriname, Kenya, Canada, Brazil, Ecuador, Colombia, Costa Rica, and the United States.

2:00 - 2:10 CST / 3:00 - 3:10 EST - Break

Presentations Session 3: GIS Data: Infrastructure and Application

Facilitator: Theresa Quill, Indiana University

2:10 - 2:50 CST / 3:10 - 3:50 EST

Revitalizing the Indiana Spatial Data Portal

Katie Chapman, Indiana University Bloomington

Over the past year, Indiana University's Research Data Services team designed and deployed a new platform for the Indiana Spatial Data Portal (ISDP) that allows IU to continue providing sustainable access to the geographic data housed in the ISDP. The portal includes data from the statewide imagery initiatives since 2005, statewide elevation datasets, NAIP imagery, regional datasets, USGS topographic maps, and Sanborn historic maps. The updates to the ISDP included replacing the existing portal with a new user interface and robust middleware, as well as workflows that ingested new data not previously available in the ISDP. The revised platform continues to rely on IU’s Intelligent Infrastructure for portal and server components and UITS Research Technologies storage and database hosting services for search and data delivery. Data on ISDP is stored in IU’s Scholarly Data Archive (SDA), a tape archive housed at IU Bloomington and Indianapolis in disaster-resistant data centers, with a redesigned file structure which makes efficient and performant use of the replicated tape system.

Geospatial Analysis of Fire Incidents in Pittsburgh using SimplyAnalytics

Charles Swartz, SimplyAnalytics

In a five-year period, house fires caused 2,620 deaths and $6.9 billion in property damage (National Fire Protection Association). It can also take just 30 seconds for a small flame to turn into a major blaze (Department of Homeland Security). Analyzing maps of fire incidents helps researchers gain valuable insights into the causes of fires and can help them understand locations that might be at higher risk for certain fire types. In this presentation, fire incident data from the Western Pennsylvania Regional Data Center (WPRDC) has been imported into SimplyAnalytics (a web-based mapping, analytics, and data visualization application with 100,000+ data variables) to use alongside demographic, health, psychographic, and consumer behavior data. Further, researchers can quickly generate reports and visualizations from the underlying data to find correlations or associations in target geographies. Utilizing data from diverse sources can aid local fire departments with community outreach and resource allocation.

2:50 - 3:00 CST / 3:50 - 4:00 EST - Break

Career Panels

3:00 - 4:00 CST (4:00 - 5:00 EST)

Facilitator: Yue Shirley Li, Purdue University

Panel 1: How to Keep Learning After Graduation: Professional Development and Continuing Education in the Spatial Workforce

For many, the goal of post-secondary education is to successfully enter the workforce in the field of your choosing. However, in an industry like GIS and Spatial where the technology is continually and rapidly changing, how do we keep our skills relevant and how do we learn in meaningful ways? Professional organizations offer many paths to answer these questions. Building networks with other spatial professionals, volunteering within an organization, and structured learning opportunities in the form of conferences and workshops, provide social outlets and skill development for continuous learners. Members of the Vanguard Cabinet discuss what it means to be involved in a professional organization, their impact in our career development, and the programs that we are developing in the Vanguard Cabinet to help other young professionals.

Panelists:

• Sid Pandey (Penn State, University of Maryland)
• Kate Berg (University of California Los Angeles, University of Michigan)
• Wanmei Liang (University of California Los Angeles)

Panel 2: URISA BIG 10 Alumni Panel

Join three members of URISA (GIS professional association) who are each Big 10 alumni sharing their geospatial career pathways. Their Big 10 educations set them each on a pathway enabling enabled each of them to gain experience with geospatial technology as it was bursting onto the scene across several disciplines during the 1990’s. Panelists were each educated at Big 10 Universities across a variety of disciplines including: environment and sustainability; forestry; rural sociology and community development and geography. Panelists will speak about their careers in private industry and consulting, academia, and local government. We will share how our educational experiences at Big 10 universities provided us with an excellent foundation for interdisciplinary exploration using geographic information systems throughout our career paths. Panelists will also emphasize the value of their involvement with geospatial professional associations including our professional development and service through URISA.

Panelists:

• Dr. Steve Steinberg, Geographic Information Officer (GIO) for the County of Los Angeles, California. (M.S., University of Michigan; Ph.D., University of Minnesota)
• Dr. Sheila Lakshmi Steinberg, Professor of GIS, Social, & Environmental Sciences. Professor of GIS, Social & Environmental Sciences, University of Massachusetts Global. (Ph.D., The Pennsylvania State University)
• Judy Colby-George, owner of Viewshed: Landscape Architecture, GIS, and Planning. (B.S., M.S., University of Wisconsin, Madison)
• Allen Ibaugh, AICP, GISP, President, Eyelock, LLC (M.S. Geography M.A. Urban Planning – University of Iowa; B.S. Psychology, B.A. Geography, Indiana University)

---

BTAA GIS Conference 2022 Planning Committee

• Karen Majewicz (Chair), Geospatial Product Manager, University of Minnesota, majew030@umn.edu
• Tara Anthony, GIS Specialist, Penn State University, tll38@psu.edu
• Caroline Kayko, Map and Geospatial Data Librarian, University of Michigan, ckayko@umich.edu
• Shirley Li, GIS Analyst, Purdue University, li1050@purdue.edu
• Laura Kane McElfresh, Cartographic Metadata Librarian, University of Minnesota, mcelf008@umn.edu
• Theresa Quill, Map and Spatial Data Librarian, Indiana University, theward@indiana.edu
• Josh Sadvari, Geospatial Information Librarian, Ohio State University, sadvari.1@osu.edu
• Nicole Scholtz, Librarian for Geospatial and Numeric Data, University of Michigan, nscholtz@umich.edu
• Amanda Tickner, GIS Specialist, Michigan State University, atickner@msu.edu