Paper type: Paper

Urban Informatics

Urban Informatics in the Science and Practice of Planning

Journal of Planning Education and Research August 2018

The vast amount of data being generated in and about cities creates both an opportunity and a dilemma for urban policymakers and planners. This paper articulates the theoretical, practical, and pedagogical foundations for the fields of urban informatics and civic analytics and outlines the challenges to effectively applying big data and computational methods to urban management, policy, and planning. It describes the state of the field, defines the range of applications in the urban context, and presents key considerations in training scientists that both acknowledge and capitalize on shifting modes of learning, working, and decision making. Situated within the ethical and moral landscape of data analytics, it articulates the knowledge and skills needed by future urban science practitioners and concludes with a discussion of data-driven problem solving in the urban context.

READ MORE

Using machine learning and small area estimation

Using machine learning and small area estimation to predict building-level municipal solid waste generation in cities

Computers, Environment and Urban Systems

Volume 70, July 2018, Pages 151-162

Municipal solid waste management represents an increasingly significant environmental, fiscal, and social challenge for cities. Understanding patterns of municipal waste generation behavior at the household and building scales is a critical component of efficient collection routing and the design of incentives to encourage recycling and composting. However, high spatial resolution estimates of building refuse and recycling have been constrained by the lack of granular data for individual properties. This paper presents a new analytical approach, which combines machine learning and small area estimation techniques, to predict weekly and daily waste generation at the building scale. Using daily collection data from 609 New York City Department of Sanitation (DSNY) sub-sections over ten years, together with detailed data on individual building attributes, neighborhood socioeconomic characteristics, weather, and selected route-level collection data, we apply gradient boosting regression trees and neural network models to estimate daily and weekly refuse and recycling tonnages for each of the more than 750,000 residential properties in the City. Following cross-validation and a two-stage spatial validation, our results indicate that our method is capable of predicting building-level waste generation with a high degree of accuracy. Our methodology has the potential to support collection truck route optimization based on expected building-level waste generation rates, and to facilitate new equitable solid waste management policies to shift behavior and divert waste from landfills based on benchmarking and peer performance comparisons.

 

Read More

A Novel Environmental Justice Indicator

Environmental justice efforts in the United States seek to provide equal protection from environmental hazards, such as air pollution, to all groups, particularly among traditionally disadvantaged populations. To accomplish this objective, the U.S. EPA has previously required states to use an environmental justice screening tool as part of air quality planning decision-making. The generally utilized approach to assess potential areas of environmental justice concern relies on static comparisons of environmental and demographic information to identify areas where minority and low income populations experience elevated environmental exposures, but does not include any additional information that may inform the trade-offs that sub-populations of varying socio-demographic groups make when choosing where to reside in cities. In order to address this limitation, job accessibility (measured by a mobility index defining the number of jobs available within a set commuting time) was developed as a novel environmental justice indicator of environmental justice priority areas at the local level. This approach is modeled using real-world data in Allegheny County, PA (USA), and identifies areas with relatively high levels of outdoor air pollution and low access to jobs. While traditional tools tend to flag the poorest neighborhoods for environmental justice concerns, this new method offers a more refined analysis, targeting populations suffering from the highest environmental burden without the associated benefits of urban living.

 

Read the Paper

Assessing Air Quality Index Awareness and Use

Background

The Mexico City Metropolitan Area has an expansive urban population and a long history of air quality management challenges. Poor air quality has been associated with adverse pulmonary and cardiac health effects, particularly among susceptible populations with underlying disease. In addition to reducing pollution concentrations, risk communication efforts that inform behavior modification have the potential to reduce public health burdens associated with air pollution.

Methods

This study investigates the utilization of Mexico’s IMECA risk communication index to inform air pollution avoidance behavior among the general population living in the Mexico City Metropolitan Area. Individuals were selected via probability sampling and surveyed by phone about their air quality index knowledge, pollution concerns, and individual behaviors.

Results

The results indicated reasonably high awareness of the air quality index (53% of respondents), with greater awareness in urban areas, among older and more educated individuals, and for those who received air quality information from a healthcare provider. Additionally, behavior modification was less influenced by index reports as it was by personal perceptions of air quality, and there was no difference in behavior modification among susceptible and non-susceptible groups.

Conclusions

Taken together, these results suggest there are opportunities to improve the public health impact of risk communication through an increased focus on susceptible populations and greater encouragement of public action in response to local air quality indices.

 

Read the Paper

Adopting Federal Approaches

Air quality regulations are intended to reduce adverse impacts to human health and the environment. Because efforts to improve air quality are not without cost, the U.S. EPA  conducts cost­‐benefit analysis to assess the health and economic impacts of air pollution abatement policies. Under the current EPA administration, it is expected that responsibility for environmental regulation will increasingly shift to state governments, some of which have little experience in conducting health and economic assessments as part of the rulemaking process. Utilizing real‐world air pollution data in conjunction with health risk estimates derived from the epidemiological literature, a scientific survey was created to assess whether localvaluations of improved air quality can be effectively estimated using federal cost‐benefit analysis methods. Results from both high and low pollution cities found that median valuations for improved air quality from survey responses were of similar magnitude to valuations used in federal cost‐benefit analysis, suggesting federal methods may be suitable for use at the local level. These methods can be augmented with information regarding local  preferences; for example, a majority of the study population was unwilling to accept reduced energy bill costs in exchange for higher pollution levels in their communities. In the absence  of state‐specific approaches, states are encouraged to adopt existing federal policy assessment tools and to fully consider the health impacts of air pollution as part of their own  regulatory review processes. 

 

Read the Paper

The Case of Win Shelters in New York City

Applications of Machine Learning Methods to Predict Readmission and Length-of-Stay for Homeless Families: The Case of Win Shelters in New York City

 

Journal of Technology in Human Services Volume 36, 2018 – Issue 1

New York City faces the challenge of an ever-increasing homeless population with almost 60,000 people currently living in city shelters. In 2015, approximately 25% of families stayed longer than nine months in a shelter, and 17% of families with children that exited a homeless shelter returned to the shelter system within 30 days of leaving. This suggests that “long-term” shelter residents and those that re-enter shelters contribute significantly to the rise of the homeless population living in city shelters and indicate systemic challenges to finding adequate permanent housing. This article focuses on our preliminary work with Win (Women-in-Need) shelters to understand the factors that predict readmission and length-of-stay of homeless families. We create a unified, comprehensive database of the homeless population being served by Win shelters, accounting for more than 6,000 homeless families. We apply logistic regression models and an unsupervised clustering algorithm to identify predictors of re-entry and long-term length-of-stay. Citizenship, age, medical conditions, employment, and history of foster care or shelter stays as a child are found to be significant predictors. The results of the K-means clustering identify three primary groups, consistent with previous typologies characterized by transitionally homeless, episodically homeless, and chronically homeless.

 

Read More

The Resilience to Emergencies and Disasters Index

The Resilience to Emergencies and Disasters Index: Applying big data to benchmark and validate neighborhood resilience capacity

Sustainable Cities and Society

Volume 36, January 2018, Pages 272-285

Resilience planning and emergency management require policymakers and agency leaders to make difficult decisions regarding which at-risk populations should be given priority in the allocation of limited resources. Our work focuses on benchmarking neighborhood resilience by developing a unified, multi-factor index of local and regional resilience capacity: the Resilience to Emergencies and Disasters Index (REDI). The strength of the REDI methodology is the integration of measures of physical, natural, and social systems – operationalized through the collection and analysis of large-scale, heterogeneous, and high resolution urban data – to classify and rank the relative resilience capacity embedded in localized urban systems. Feature selection methodologies are discussed to justify the selection of included indicator variables. Hurricane Sandy is used to validate the REDI scores by measuring the recovery periods for neighborhoods directly impacted by the storm. Using over 12,000,000 records for New York City’s 311 service request system, we develop a proxy for neighborhood activity, both pre- and post-event. Hurricane Sandy had a significant and immediate impact on neighborhoods classified as least resilient based on the calculated REDI scores, while the most resilient neighborhoods were shown to better withstand disruption to normal activity patterns and more quickly recover to pre-event functional capacity.

 

Read More

Urban phenology

Urban phenology: Toward a real-time census of the city using Wi-Fi data

Computers, Environment and Urban Systems

Volume 64, July 2017, Pages 144-153

New streams of data are being generated by a range of in-situ instrumentation, mobile sensing, and social media that can be integrated and analyzed to better understand urban activity and mobility patterns. While several studies have focused on understanding flows of people throughout a city, these data can also be used to create a more spatially and temporally granular picture of local population, and to forecast localized population given some exogenous environmental or physical conditions. Effectively modeling population dynamics at high spatial and temporal resolutions would have significant implications for city operations and policy, strategic long-term planning processes, emergency response and management, and public health.

This paper develops a real-time census of the city using Wi-Fi data to explore urban phenology as a function of localized population dynamics. Using Wi-Fi probe and connection data accounting for more than 20,000,000 data points for the year 2015 from New York City’s Lower Manhattan neighborhood – combined with correlative data from the U.S. Census American Community Survey, the Longitudinal Employer-Household Dynamics survey, and New York City administrative records – we present a model to create real-time population estimates classified by residents, workers, and visitors/tourists in a given neighborhood and localized to a block or geolocation proximate to a Wi-Fi access point. The results indicate that the approach has merit: we estimate intra-day, hourly worker and resident population counts within 5% of survey validation data. Our building-level test case demonstrates similar accuracy, estimating worker population to within 1% of the reported building occupancy.

 

Read More

Accuracy of Quantification of Risk

Accuracy of quantification of risk using a single-pollutant Air Quality Index

Journal of Exposure Science and Environmental Epidemiology, 2017 Jan; 27(1):24-32.

 doi: 10.1038/jes.2015.43

Health risks associated with short-term exposure to ambient air pollution are communicated to the public by the US EPA through the Air Quality Index (AQI), but it remains unclear whether the current regulatory-based, single-pollutant AQI fully represents the actual risks of air pollution-related illness. The objective of this study is to quantify cardiovascular hospital admissions attributable to PM2.5 at each AQI category. Based on National Ambient Air Quality Standards (NAAQS), the highest AQI value among criteria pollutants (driver pollutant) is reported daily. We investigated excess cardiovascular hospital admissions attributable to fine particulate matter (PM2.5) exposure from 2000 to 2010 in Bronx, Erie, Queens, and Suffolk counties of New York. Daily total, unscheduled cardiovascular hospital admissions (principal diagnosis) for individuals aged 20-99 years, concentration-response functions for PM2.5, and estimated quarterly effective daily concentrations were used to calculate excess cardiovascular hospital admissions when PM2.5 was reported as the driver pollutant and when PM2.5 was not reported as the driver pollutant at each AQI category. A higher proportion of excess hospital admissions attributable to PM2.5 occurred when PM2.5 was the driver pollutant (i.e., ~70% in Bronx County). The majority of excess hospital admissions (i.e., >90% in Bronx County) occurred when the AQI was <100 ("good" or "moderate" level of health concern) regardless of whether PM2.5 was the driver pollutant. During the warm season (April-September), greater excess admissions in Suffolk County occurred when PM2.5 was not the AQI driver pollutant. These results indicate that a single-pollutant index may inadequately communicate the adverse health risks associated with air pollution.

 

Read the Paper