I'm a data scientist at LWA and spend most of my time in
R automating ETL pipelines for sensor networks 📡, building Shiny Apps and dashboards 🖥, designing approaches with spatial statistics and hydrologic models, and generally wrangling lots and lots of data.
I have a PhD in Hydrology and my dissertation is titled ‘Emerging consequences of regional-scale aquifer depletion: data-driven and numerical models of well failure, basin salinization, and contaminant transport’ (my exit seminar can be viewed here). Early in my PhD, I found that I really enjoyed data science and programming, and I used these years to sharpen those skills. My published research includes NLP and network analysis, spatial statistics, and physical modeling of 3D, subsurface contaminant transport.
I'm an #rstats nerd and automation/reproducibility fanatic. My favorite tools include tidyverse (
RMarkdown (for dashbaords/reporting),
leaflet (for spatial data), and
DBI for databases. A few projects I'm proud of include an
R package to query water quality data 📦, R data science curriculum 📚, a dashboard that makes millions of water quality observations understandable 📈, and a model that predicts the risk of wells going dry 💧 funded by Microsoft's AI for Earth Grant.
Before my PhD, I taught environmental science to middle and high school students in Yosemite and the Marin Headlands for the educational nonprofit NatureBridge. I spent summers leading expeditions in the wilderness, and in Thailand for National Geographic Student Expeditions.
In my free time, I enjoy anything that puts me in a flow state: alpine climbing 🧗🏼, running 🏃♂️, surfing 🏄♂️, tinkering on bikes 🚴♂️, reading 📚, playing guitar 🎸, and cooking 🧑🍳.
PhD in Hydrogeology, 2020
University of California Davis
BA in Integrative Biology, minor in Conflict Resolution, 2011
University of California Berkeley
Domestic well failure prediction and cost estimates in critically overdrafted basins.
Automated water quality reports for > 3,000 California public water systems. 🏆 Winner 2019 California Water Data Challenge.
Exploratory data analysis of California's Online State Well Completion Report Database.
Using random forests, boosting, LDA, and QDA with variable probability thresholds for global landslide classification.