iiinova — where your digital self keeps evolving

Daniel Horton @danethan.bsky.social · 06/10/2026

We look at many precip products in the context of post-wildfire debris flows in southern CA. We have some (5) recommendations, but chief amongst them is the first:

000

Daniel Horton @danethan.bsky.social · 06/10/2026

Precip products can be radar-based, satellite-based, gauge-based, physics-based, gauge-/radar-/satellite-based, radar-/gauge-/physics-based, hindcast, forecast, etc.

Summary of gridded quantitative precipitation estimate (QPE) products included in this study

110

Daniel Horton @danethan.bsky.social · 26/08/2026

Turns out, heat has all sorts of negative health impacts...

Stage 1 screening results across 1803 diagnosis categories. each point represents an icd-10 diagnosis code; the y axis shows −log10 of the Bh-adjusted P value from a quasi-Poisson model adjusted for calendar time, day of week, holidays, and same-day cook county ePA monitor–based daily PM2.5. the red dashed line marks Bh- adjusted P = 0.10. labels identify the 44 codes meeting all stage 1 criteria. text labels are displayed for all codes that met all screening criteria: (i) a positive association between ed/Uc visit counts and the 3-day moving average of daily maximum temperature, (ii) Bh-adjusted P < 0.1, (iii) ≥30% of ed/Uc visits occurring on days above the 70th percentile of the temperature distribution (29.74°c), and (iv) at least 100 counts of the diagnosis code recorded during warm-season (May to September) in the ed/ Uc systems. chapter 19 and 20 (injury/poisoning and external cause) labels are shown in inset boxes to avoid overplotting.

Cumulative ORs at lags 0 to 3 from stage 2 DLNMs. Points show cumulative ORs at 33.67°c relative to the diagnosis-specific constrained MRt; horizontal bars show 95% cis. Panel presents 10 disease-related diagnoses. Reported codes had cumulative lag 0-to-3 lower 95% ci > 1 and ePv ≥ 10. the vertical dashed line marks OR = 1.

010

Daniel Horton @danethan.bsky.social · 13/04/2026

Elated to announce a banner NSF GRFP year for my lab group: Grad student Anna Oehlerking, ugrad researcher Kyan Shlipak (MIT bound), and incoming grad student Maggie Gonzalez were all selected. Proud of these amazing young scientists, and honored to work with them.

140

Daniel Horton @danethan.bsky.social · 25/03/2026

🚨🧪🚨New research from Dr. Peter Graffy, just published in Journal of Biomedical Informatics: Heat and hearts: An exposure-anchored computational phenotyping framework for assessing cardiovascular vulnerability during extreme heat. 🔥+❤️=❤️‍🔥&🥵 doi.org/10.1016/j.jb...

This is the Graphical Abstract of Heat and hearts: An exposure-anchored computational phenotyping framework for assessing cardiovascular vulnerability during extreme heat

100

Daniel Horton @danethan.bsky.social · 29/07/2025

The methods presented in this study are innovative and interesting, but there's still lots of work to be done to advance the methodology, much of which is outlined in the Discussion section, so please have a look :)

Regional landslide hydrometeorology framework

000

Daniel Horton @danethan.bsky.social · 29/07/2025

In this study she applies these methods to a California case study, i.e., the December 2022 to January 2023 parade of 9 consecutive landfalling atmospheric rivers that caused 100s of landslides across the state.

100

Daniel Horton @danethan.bsky.social · 29/07/2025

Chuxuan's thesis, funded by an NSF PREEVENTS award, explores methods to integrate regional hydrological modeling into land surface hazard research. Chuxuan developed a water balance framework that provides a means to assess hydrological conditions relevant to landslide events over a regional domain.

Landfall locations of atmospheric rivers, location of reported landslides, and precipitation info

100

Daniel Horton @danethan.bsky.social · 06/07/2025

Observed Changes in Daily Precipitation Intensity in the United States -> doi.org/10.1029/2022...

Change in mean wet day intensity; Continental U.S.

Change in frequency across intensities; Continental U.S.

120

Daniel Horton @danethan.bsky.social · 28/02/2025

Getting transport emissions right, particularly at high spatial resolutions, is consequential for accurately simulating air pollution and its impacts, including downstream distributional impacts, such as assessments of exposure, health, and equity.

Relative NOx emission disparities for racial and ethnic subgroups, calculated using the FIVE, NEMO, LADCO-NU, and UVM data sets for the Greater Chicago Region. Disparities are expressed as the ratio of subgroup population-weighted emissions to the domain-wide population-weighted average.

100

Daniel Horton @danethan.bsky.social · 28/02/2025

As learned via efforts like CMIP, 1 of our best means for improving models is to compare output from different groups. Here Victoria & team compare 4 hi-res transport emission datasets over Chicagoland, (FIVE, NEMO, LADCO, & UVM) identifying similarities, differences, & areas for improvement.

Spatial distribution of simulated NOx emission differences from the four-data set mean (data set minus mean) across the Greater Chicago region: (a) FIVE, (b) NEMO, (c) LADCO-NU, and (d) UVM at 1 km2 spatial resolution. Insets are centered on the “Loop”, and segments of I- 294 west, and south of downtown Chicago.

100

Daniel Horton @danethan.bsky.social · 04/12/2024

Among many interesting things, he found an absolute unit of a redwood log jam (subplot D), which spurred grain size recovery just 2 years post-fire.

Lidar differencing and giant pile of wood

000

Daniel Horton @danethan.bsky.social · 04/12/2024

Telly used pre- and post-fire structure from motion imagery to track stream morphology and sediment changes.

Subsets of structure from motion orthoimages from 2016, 2021, and 2022 surveys in (a) Big Creek and (b) Devil's Creek, CA representing pre- and post-fire time points.

110

Daniel Horton @danethan.bsky.social · 17/04/2024

Attending #EGU24? If so, check out Sara Camilleri's presentation in #EDI Session "Incorporating equity, gender, health and other co-benefits into NEXUS and IAM research" on Thursday (18 Apr) at 14:55 CEST in Room 2.17 meetingorganizer.copernicus.org/EGU24/EGU24-...

000

Daniel Horton @danethan.bsky.social · 26/03/2024

Joint AGU/AMS Climate and Health Showcase on April 3, Wed, from 10am to 4:45pm, EST! Don't miss keynote speakers: Dr. Yang Liu and Kris Ebi.

000

Daniel Horton @danethan.bsky.social · 07/11/2023

"Across CONUS, we estimate that the NO2-attributable all-cause mortality is ∼170,850 (95% confidence interval: 43,970, 251,330) premature deaths per yr with large variability across census tracts and within individual cities."

100

Daniel Horton @danethan.bsky.social · 22/09/2023

Stacy used a clustering method to determine high-confidence hotspots, but also leveraged dataset disagreements to highlight areas for model improvement and targeted observations.

100

Daniel Horton @danethan.bsky.social · 22/09/2023

This work sprung from Stacy's summer internship at Microsoft Research. Her query: with the advent of multiple high spatial resolution state-of-the-science air quality datasets, where are we confident that NO2 pollution is bad & what can we learn when the datasets disagree?

100

Daniel Horton @danethan.bsky.social · 15/09/2023

Bottom line: EVs are not the solution to all our climate, air quality, transport, and EJ problems…but they’re much better than ICE vehicles.

020

Daniel Horton @danethan.bsky.social · 15/09/2023

On a per vehicle basis Visa et al can be compared to Camilleri et al's eHDV study to show the outsized impact of high-emitting heavy-duty vehicles, BUT sheer # of passenger cars/trucks suggests incentives needed for ALL classes; or better yet alternative/active transport options

100

Daniel Horton @danethan.bsky.social · 15/09/2023

We find that EV adoption decreases GHG emissions (even w/ a fossil fueled battery charging grid), decreases NO2 & PM pollution, and increases O3 pollution – but that health benefits are overall very positive, and disproportionately so in marginalized communities. An EJ win.

110