This project was a collaboration between the Sustaining Urban Places Research Lab, Willamette Partnership, and Clean Water Services. The aim of the project was to assess the reduction in vehicular emissions due to a massive tree planting campaign led by Clean Water Services. Alongside multi-year canopy growth scenarios and pollution mitigation models, I created a high-resolution (1ft) dasymetric population map for the tri-county area which allowed for direct health impacts to be analyzed.
This data was created in collaboration with Oregon Metro and published in the Regional Land Information System (RLIS). The classification has a high overall accuracy (88%; Kappa = 0.75) and a spatial resolution of 1m². In order to attain such a high accuracy, we developed an ensemble machine learning system which incorporated techniques such as Neural Networks, Random Forests, and Support Vector Machines. The ensable was trained on ~900 variables which included trigonometric crown measures, spectral indices, and localized terrain characteristics. In addition to the classification, this project also identified ~13.4 million trees in the Portland metro area.
This project seeks to provide information to public health and planning practitioners as to the resiliency of specific sub-neighborhoods in Portland, Oregon in terms of environmental health detriments. The tool allows the user to specify weights of multiple variables that reduce an individual's ability to cope to environmental stressors, which updates the map with locations most pertinent to the user. This socio-environmental analysis was developed through a multi-stage interagency multi-criteria decision making exercise. The following agencies, bureaus, and organizations were involved in the process:
The Plantability tool is part of a substantial effort by Portland Parks and Recreation Urban Forestry and the Sustaining Urban Places Research Lab to understand the social, environmental, and technical aspects involved in increasing the overall canopy cover within the city. The development of this tool included collaboration with multiple community groups and public agencies and bureaus.
Quantifying and mapping spatial patterns of extreme intra-urban heat variability was my main focus during my time as a Geospatial Research Analyst at the SUPR Lab. As I transitioned into graduate school, I continued researching better ways to identify areas of cities where people are most at risk. A major difference between the method used in my research and common methods were the in-situ (or, "on the ground") measurements taken in the field. These measurements allow for high-resolution and high-accuracy machine-learning models to be created, as opposed to low-resolution and low-accuracy satellite data employed in other studies. This research eventually moved beyond Portland, and I have worked with local governments, non-profits, and universities to accomplish field data collection in:
Throughout my time at the Sustaining Urban Places Research Lab and during my Masters of Urban Studies research, lidar has been a constant focus. Using a combination of proprietary and open-source tools, I have processed several terabytes of 3D pointcloud data and converted it into useable vector and raster GIS datasets.
Through extensive remote sensing and lidar exercises, I have created unique descriptors of Portland, Oregon's urban forests. Throughout presentations with various government and community groups, I have shared my findings in hopes of raising awareness of the magnificent trees that surround us.
The Canopy Analytics tool was developed in order to identify large trees (which contribute greatly to reductions in temperatures in air pollution) that are most in need of preservation. The tool allows users to specify heights, building distances, and other factors that the City of Portland deems crucial in determining the risk a tree has of being removed.
Davis, A. C., Voelkel, J. L., Remmers, C. L., Adams, J. L., & McGlynn, E. A. (2023). Comparing Kaiser Permanente Members to the General Population: Implications for Generalizability of Research. The Permanente Journal, 1-12.
Friedman, R. S.C., Carpenter, D. M., Saver, J. M., McDermott, S. C., Voelkel, J. (2022). Telemedicine Familiarity and Post-Disaster Utilization of Emergency and Hospital Services for Ambulatory Care Sensitive Conditions. American Journal of Preventive Medicine, 63(1), e1.
Shandas, V., & Voelkel, J., Williams, J., Hoffman, J. (2019). Integrating Satellite and Ground Measurements for Predicting Locations of Extreme Urban Heat. Climate, 7(1), 5.
Voelkel, J., & Hellman, D., Sakuma, R., Shandas, V. (2018). Assessing vulnerability to urban heat: a study of disproportionate heat exposure and access to refuge by socio-demographic status in Portland, Oregon. International journal of environmental research and public health, 15(4), 640.
Voelkel, J., & Shandas, V. (2017). Towards Systematic Prediction of Urban Heat Islands: Grounding Measurements, Assessing Modeling Techniques. Climate, 5(2), 41.
Ferwati, S., Skelhorn, C., Shandas, V., Voelkel, J., Shawish, A., & Ghanim, M. (2017). Analysis of urban heat in a corridor environment–The case of Doha, Qatar. Urban Climate.
Shandas, V., van Diepen, A., Voelkel, J., & Rao, M. (2016). Coproducing Resilience through Understanding Vulnerability. Building a Climate Resilient Economy and Society – Challenges and Opportunities.
Voelkel, J., Shandas, V., & Haggerty, B. (2016). Developing High-Resolution Descriptions of Urban Heat Islands: A Public Health Imperative. Preventing chronic disease, 13.
Shandas, V., Voelkel, J., Rao, M., & George, L. (2016). Integrating High-Resolution Datasets to Target Mitigation Efforts for Improving Air Quality and Public Health in Urban Neighborhoods. International journal of environmental research and public health, 13(8), 790.
Banis, D., & Shobe, H. (2015). Portlandness: A Cultural Atlas. Sasquatch Books. [Contributing Cartographer].
Custom tool creation, data creation and manipulation, advanced raster processing and analysis, automation, remote sensing (air- and space-borne, including hyperspectral and thermal), lidar analysis, webmapping (Leaflet, OpenLayers 3), database design and normalization (ESRI File Geodatabase, PostgreSQL/PostGIS)
Explanatory and predictive, Logistic Regression, Machine learning (Support Vector Machines, Random Forests, and Neural Networks), accuracy assessments, spatial statistics (including geographically weighted regression).
Microsoft Windows 7 and 10, OSX, UNIX/Linux
R (as both a utilitarian language and a statistics application), Python (with comprehensive ArcPy knowledge), Bash / Command Line, HTML, CSS, JavaScript, SQL, IDL (ENVI lidar API)
R/Rstudio, ArcGIS Suite, QGIS, GRASS GIS, SAGA GIS, Idrisi GIS, Google Earth Engine, ENVI, ENVI lidar, ENVI IDL, IBM SPSS, Microsoft Office / Libreoffice, PgAdmin, Adobe Creative Suite, open source graphics software (GIMP, InkScape), CMD/Terminal