THE UNIVERSITY OF CHICAGO
CSDS Study Group Presentations: 2023
Roger Bivand, Professor Emeritus
Jan 17 , 2023
PROJ, GEOS and GDAL have changed a lot in the last 5 years. In addition, M1 has arrived, as has UCRT for Windows, causing a lot of churn in R-spatial package maintenance. Further, three packages maintained by me: rgdal, rgeos, and maptools, are to retire during 2023. There are more modern alternatives and developer effort is better applied to supporting them. The raster package already uses terra for access to shared FOSS foundations. The sp package is being re-based to use sf, so it will remain. The talk will report on progress so far, the current status of the evolution project, and prospects for completion this year.
Roger Bivand, Professor Emeritus
Jan 19 , 2023
The FOSS ecosystem is very dependent on the PROJ library, which has changed a great deal since PROJ.4. PROJ 5 introduced transformation pipelines, PROJ 6 an SQLite database of CRS, transformations, etc., and upgraded CRS representation from the legacy PROJ.4 string to WKT2-2019. PROJ 7 introduced a content download network for accessing transformation grids. We are now at PROJ 9.1.1. GDAL 3 requires PROJ 6 or greater, and is now tightly integrated with PROJ. In addition, modern vector file formats such as GeoPackage assume a basic geographical CRS where none is given, so we need to innovate to express a missing planar CRS.
Dash Young-Saver
Jan 24 , 2023
Skew The Script is a nonprofit that provides free, socially relevant math lessons to 400,000 high school students across the country. We create lessons on relevant topics (e.g. food deserts, gerrymandering, online dating, sports) to make math class compelling, especially for students from underserved backgrounds. We need your help! We're making a lesson on sampling methods and income segregation, which requires an interactive map of incomes in every major metro area in the US. The problem: We lack the skills to make our current map workable for students and teachers. Lend us your ear, eyes, and expertise during this live feedback session, to help us make an impact in thousands of high school classrooms across the country.
Wataru Morioka
Jan 31 , 2023
Capturing spatial co-location patterns—subsets of two or more types of events that are geographically close—is one of the primary interests in spatial analysis. For example, in many central districts in cities across the world, different types of stores form clusters to take advantage of spatial agglomeration. To precisely analyze the micro-scale co-location in downtown area, we developed a set of extended statistical methods named network dual K function, which is based on Ripley's K function. The proposed methods were applied to various types of stores in trendy districts in Tokyo to demonstrate the usefulness of the method.
Marynia Kolak
Feb 7 , 2023
Accessing evidence-based medications for opioid use disorder (MOUD) and harm reducing resources is critical to saving lives and healing communities, yet little is known about how access to these resources is best measured. How far is too far, and how does travel behavior (eg. walking versus driving) and medication type (eg. methadone versus buprenoprhine) impact outcomes? Understanding transit behaviors and resource variability across varying degrees of urbanicity are crucial to modeling MOUD needs, though persistent challenges of stigma of resource usage present a complex challenge for access researchers in this space. Our goal is to develop a systematic analysis of how researchers measure access for MOUD and relevant resources, calculate and compare spatial access metrics, and identify optimal thresholds of success. A multidimensional view of access, with more refined spatial measures incorporated, may better identify places or characterize community profiles for intervention and resources.
Niall Newsham
Feb 14 , 2023
The goal of his presentation will be to get feedback about research that is currently under development:
Population decline is becoming widespread across Europe, with the reversal of longstanding population growth imminent. Though population change is inherently a localised process, forecasts are often only produced at large spatial scales, and are insufficient to prepare for unprecedented change. Recent developments in machine-learning based forecasting methodologies present an exciting opportunity to understand demographic futures. Making use of WorldPop gridded population estimate data, we aim to develop a machine-learning model to create high-resolution population decline estimations across Europe. Though the exact methods are yet to be decided, we propose a recurrent neural network (RNN) approach to create a predictive model of European population decline. Feedback is particularly welcomed!
Luc Anselin & Pedro Amaral
Feb 21 , 2023
Whereas the estimation of spatial regime regressions is well understood, the delineation of the regimes themselves remains a topic of active interest. In this paper, we outline a heuristic to determine the spatial regimes endogenously, as an extension of the well-known SKATER algorithm for spatially constrained clustering. This guarantees that the resulting regimes consist of contiguous observations. We outline the method and apply it in the context of the determination of housing submarkets, which is represented by a rich literature in applied spatial econometrics. We compare the estimation of a hedonic house price model using the endogenous spatial regimes approach to a range of more traditional methods, including pooled regression, the use of administrative districts, data-driven regimes based on a-spatial and spatial clustering of explanatory variables, and finite mixture regression. We evaluate the results in terms of fit and assess the trade-offs between the spatial and a-spatial approaches.
Carmen Villa Llera
March 28 , 2023
Youth centres are publicly funded spaces for young people to spend time after-school. They usually target lower income populations who may not have the means for other paid leisure activities and often lack positive role models at home. Since 2010 nearly 44% of all youth centres that were open in London had to close due to austerity. I have created a database with the location of youth centres operating in London in 2010-2019 and gathered information on closing dates. I match this spatially to administrative records from the London Metropolitan Police with information on offenders age, home address, and time of offence. I assess how changes in the distance to the nearest youth centre affect the number of crimes committed by residents, comparing areas where the distance to the nearest youth centre increases vs areas in which this distance remains the same. The preliminary results show that a 1% increase in the distance to the nearest open youth centre increases youth-crimes committed in after-school hours by 7%.
Luisa Fernanda Eusse Villa
April 4 , 2023
Over the last two decades, numerous studies have revealed that people's preferences toward non-market and market goods and services are complex and vary across space. Identifying the sources of such heterogeneity in preferences provides decision-makers with valuable information to develop better-informed and designed policies that target programs more efficiently and consistently with consumer preferences. In my talk, I will discuss my doctorate research on modelling spatial heterogeneity in consumer preferences, which explores how the spatio-temporal context affects willingness to pay estimates based on some of my previous work. I will also present my current research, which is based on recently collected data and is a work in progress, thus, the goal of this presentation will be to get feedback on appropriate spatial methods and data analysis. My current research seeks to analyze if the preferences of Italian consumers for pecan nuts vary across the national territory by assessing the effect of the spatial context on their preferences.
Dr. Nick Mader
April 11 , 2023
Access to childcare is a critical support for families in engaging in work and for supporting development of young children. But while many federal, state, and local funds are targeted to making childcare affordable, and thus accessible, many families are unaware of or unable to take these funds. State and local policymakers have the means to improve take-up, but lack a reliable data on where to find subsidy-eligible families that are not yet engaged, as Census data reports variously lack key measures, lag significantly behind economic trends that impact eligibility, or are too geographically aggregate. This project represents both new statistical methods and turn-key production code to provide policymakers with the information needed to drive concrete policy efforts to engage families with benefits for which they are eligible.
Dr. Claire Kelling
April 18 , 2023
Newly available point-level datasets allow us to relate police use of force to other events describing police behavior. Current methods for relating two point processes typically rely on the spatial aggregation of one of the two point processes. We investigate new methods that build upon shared component models and case-control methods to retain the point-level nature of both point processes while characterizing the relationship between them. We find that the shared component approach is particularly useful in flexibly relating two point processes, and we illustrate this flexibility in simulated examples and an application to Chicago policing data.
Johannes Moser
April 25 , 2023
The Economic Geography literature has identified High-Speed Rail (HSR) to be a driver of economic growth for connected regions. One hypothesis is that particular types of industries, often so-called knowledge-intensive industries, are especially attracted to HSR. Knowledge-intensive firms can bring highly qualified workers and tax revenues to cities and thereby enhance their attractiveness. Germany has historically favored high regional accessibility by rail and also built a HSR network. This research project focuses on a selection of eight case studies in Germany to examine whether HSR stations indeed attract knowledge-intensive industries in their direct surroundings, for which a rich and georeferenced data set detailing firm locations is used. Tentative results indicate the absence of strong immediate effects. This is work in progress -- feedback is welcome!
Egor Kotov
May 2 , 2023
Income inequality and socio-economic spatial segregation have increased worldwide, including in many European cities, over the recent decades. Segregation can lead to an increased burden on already disadvantaged groups by overexposing them to air and noise pollution as well as other negative environmental effects. One such threat in Spain is the spread of tiger mosquitoes that can carry Dengue, Zika, and Chikungunya and they also present a significant nuisance. Species distribution models consistently show that most of the Spanish population will become even more affected in the future, but these effects may not distribute equally.
This research aims to reveal spatially segregated groups in Spain who are disproportionately exposed to mosquito-related risks. Mosquito exposure inequalities are evaluated in densely populated Spanish functional urban areas at high spatial resolution. The activity space approach allows us to go beyond the simplified understanding of socio-spatial segregation only through the residential location and evaluate the segregation based on the range and available urban amenities, as well as the overlap of activity space of groups with different incomes. Using human mobility patterns based on call-detail records enriched within income, gender, and age data and crowdsourced mosquito reports from the Mosquito Alert app, we compare mosquito encounter probabilities with activity space by socio-economic groups at the census district level.
Dr. Hamza Rarou
May 9 , 2023
This work aims to describe and examine the spatial formation and evolution of the global automotive assembly footprint. I construct a theoretical framework that combines insights from the evolutionary economic geography and international business literatures and develop several hypotheses that are tested with Cox proportional hazard regression models. The analysis is based on a unique dataset of equity-owned plant entries between 1899 and 2018. Results largely confirm time of entry, agglomeration economies, liability of foreignness, joint-venture entry mode, economic development status, and labor law regulations matter for survival rate of plants. However no statistically significant effect was found for prior host country experience. Hazard rates of higher-equity plants in developed and developing economies did not differ.
Evgeny Noi
May 16 , 2023
Exploratory data analysis tools designed to measure global and local spatial autocorrelation (e.g. Moran’s I statistic) have become standard in modern GIS software. However, there has been little development in amending these tools for visualization and analysis of patterns captured in spatio-temporal data. We design and implement an exploratory mapping tool, VASA (Visual Analysis for Spatial Association), that streamlines analytical pipelines in assessing spatio-temporal structure of data and enables enhanced visual display of the patterns captured in data. Specifically, VASA applies a set of cartographic visual variables to map local measures of spatial autocorrelation and helps delineate micro and macro trends in space-time processes. Two visual displays are presented: recency and consistency map and line-scatter plots. The former combines spatial and temporal data view of local clusters, while the latter drills down on the temporal trends of the phenomena. As a case study, we demonstrate the usability of VASA for the investigation of mobility patterns in response to the COVID-19 pandemic throughout 2020 in the United States. Using daily county-level and grid-level mobility metrics obtained from three different sources (SafeGraph, Cuebiq, and Mapbox), we demonstrate cartographic functionality of VASA for a swift exploratory analysis and comparison of mobility trends at different regional scales.
Based on these articles: