Lightnews — Scholar-powered news

Matthias Sprenger @matthiasprenger.bsky.social · 1d

This work is part of the @cuahsi.bsky.social & USGS Powell Center supported synthesis working group AI4PF on "Using a network of networks for high-frequency multi-depth soil moisture observations to infer spatial & temporal drivers of subsurface preferential flow" & we are grateful for the support!

1 3

Matthias Sprenger @matthiasprenger.bsky.social · 1d

Given the ubiquity of PF & potential higher rainfall intensity & NPP due to climate change, PF patterns are expected to change. This will increase uncertainty in predictions of groundwater recharge & water quality, which highlights need to incorporate PF into hydrological & biogeochemical models.

1

Matthias Sprenger @matthiasprenger.bsky.social · 1d

We utilized two established detection approaches—the Velocity Threshold (VT) and Non-Sequential Response (NSR) methods. Although the absolute number of detected PF events varied, both methods consistently showed strong and similar relationships between PF occurrence and its key drivers.

1

Matthias Sprenger @matthiasprenger.bsky.social · 1d

PF was more likely where antecedent soil moisture variability was low, which was found to be more important than the mean antecedent soil moisture conditions.
PF also generally increased with higher Net Primary Productivity (NPP) & in more humid climates (macropore formation through root growth).

Partial dependence plots of selected variables from random forest models of Preferential flow (PF) occurrence using velocity threshold (VT) and non-sequential response (NSR) criteria. The partial dependence of PF occurrence was based on PF derived by the VT (purple) and NSR (orange) method, respectively. In addition to the VT method based on the threshold as the maximum depth-weighted Ksat (solid dark purple line), we also show results for the median of the depth weighted Ksat (light dashed purple line), and the median of depth weighted Ksat + S.D. for each sensor (dashed purple line) to reflect uncertainties. Variables were chosen based on feature importance (Figure 2): (a) rainfall intensity, (b) clay content, (c) coefficient of variation in antecedent soil moisture, (d) mean antecedent soil moisture, (e) aridity index, and (f) net primary productivity (NPP). We defined classes based on definitions by the American Meteorological Association (AMS, 2024) for rainfall classes, UNEP for aridity (Middleton et al., 1997), Saugier et al. (2001) for NPP, and the USDA Soil Science Division Staff (2017) for soil texture classes, respectively. NPP classes do not mean specific vegetation types.

1

Matthias Sprenger @matthiasprenger.bsky.social · 1d

We identified the primary drivers using Random Forest models. Rainfall intensity, soil texture, & antecedent soil moisture emerged as critical factors:
PF likelihood steeply increased between 5 & 12 mm/h, demonstrating a critical threshold-like behavior.
PF was more likely with higher clay content.

Feature importance and the performance of random forest models in estimating Preferential flow (PF) occurrence. (a) Permutation feature importance from the random forest model at site level (40 National Ecological Observatory Network (NEON) sites) for velocity threshold (VT). (b) The model predicted and observed PF probability at each NEON site for training and testing data based on VT. (c) Same as a, but generated for non-sequential response (NSR) based PF probability. (d) Same as b, but based on NSR data. Note: Fitted probability of PF: the probability predicted by random forest model; Probability of PF: the probability derived using NEON soil moisture and precipitation data sets.

1

Matthias Sprenger @matthiasprenger.bsky.social · 1d

We used data from approximately 1500 sensors at 40 NEON sites across the USA. Our findings show that PF is ubiquitous and occurred at all sites studied. Specifically, sites experienced PF in up to 60% of all rainfall events >2 mm. This confirms PF is a pervasive feature of hydrological systems.

The distribution of the occurrence of Preferential flow (PF). (a) A map of PF occurrence across 40 National Ecological Observatory Network (NEON) sites. The probability of PF is calculated as a ratio of the number of precipitation events (≥2 mm) that triggered PF to the total number of precipitation events (≥2 mm) for which soil moisture data was available between 2016 and 2022. The left side of each circle contains the PF probability for the velocity threshold (VT) criterion, while the right side represents the result from the non-sequential response (NSR) criterion. Shown is the continental USA and Puerto Rico with gray boundaries representing 17 ecoregions. (b) Violin plots that show the distribution of the probability of PF based on VT and NSR criteria. The black boxes inside represent the 25th (q25) and 75th (q75) percentiles, with the median shown as a white point. The top and bottom fences show the difference of q25 and 1.5 times interquartile range (q75−q25) and sum of q75 and 1.5 times interquartile range, respectively. (c) The probability of PF at each NEON site based on the VT criteria against probability of PF based on NSR and stratified by peak rainfall intensity.

1

Matthias Sprenger @matthiasprenger.bsky.social · 1d

PF describes the rapid movement of water that bypasses a large fraction of the soil matrix. Thus, PF is critically influencing groundwater recharge and contaminant transport. In this study, we used the velocity threshold method and the non sequential response method doi.org/10.1002/vzj2...

Two graphs showing the soil moisture response to rainfall. One graph shows how different sensors in different soil depths respond at different times, which allows to infer preferential flow via the non sequential response analyses. The other figure shows the time difference between rainfall onset and soil moisture response onset, which is used for the verlocity threshold method to infer preferential flow. The figures are from Nimmo et al. (2025, doi https://doi.org/10.1002/vzj2.70017)

1

Matthias Sprenger @matthiasprenger.bsky.social · 1d

Our new research analyzing high-frequency soil moisture data from 40 NEON sites across 17 US ecoregions reveals the widespread occurrence of Preferential Flow (PF) & its drivers across continental scale environmental gradients. Published in AGU's Geophysical Research Letters
doi.org/10.1029/2025...

Ubiquity and Causes of Soil Water Preferential Flow Across 17 Ecoregions

Preferential flow (PF) is ubiquitous across the USA and occurs in up to 60% of all rainfall events ≥2 mm Rainfall intensity, soil texture, and antecedent soil moisture emerge as critical in gener...

doi.org

2 6 19

Reposted by Matthias Sprenger

AGU Hydrology H3S @agu-h3s.bsky.social · 3d

Keep those votes rolling in! The second head-to-head for Round 1's snow-dominated catchments is...East River vs Oldman!

Scroll to the thread below for facts on these catchments! And VOTE VOTE VOTE!!!
www.agu-h3s.org/rivercup

1 3 6

Reposted by Matthias Sprenger

Laura Helmuth @laurahelmuth.bsky.social · 3d

Six of the nine Nobel Prize winners this year work in the U.S.
Three of the six were born outside the U.S., which is the pattern most years. No country has benefited more from welcoming immigrants from around the world.
www.nobelprize.org

The official website of the Nobel Prize - NobelPrize.org

The Nobel Prize rewards science, humanism and peace efforts. This is one of the central concepts in the will of Alfred Nobel, and it also permeates the outreach activities that have been developed for...

www.nobelprize.org

27 450 1.2K

Reposted by Matthias Sprenger

JP Gannon @jpgannon.bsky.social · 4d

More about the app Paolo referenced, and that I am excited about (!) here: onlinelibrary.wiley.com/doi/10.1002/...

Interactive Learning in Hydrological Modelling With a Web‐Based Tool

We introduce a web-based tool to explore hydrological modeling in classrooms. The app can be run in any internet browser on a variety of devices. The app and code are available for free. We include m...

onlinelibrary.wiley.com

1 3

Matthias Sprenger @matthiasprenger.bsky.social · 3d

I was actually thinking of JP's great work when I saw Gemini for model development in the classroom. You can see the HTML code in the browser, yes. I was mainly thinking about overcoming the barrier of coding in exploring model parameters, input - output relationships, & process feedbacks.

1

Matthias Sprenger @matthiasprenger.bsky.social · 4d

As a new prof, I am exploring approachable ways to teach about catchment hydrology modeling. A lack of students'coding skills are limiting the opportunities. Google Gemini with Canvas has the potential to code simulators based on text input. See screen recording of quick example. What do you think?

1 3

Reposted by Matthias Sprenger

CUAHSI @cuahsi.bsky.social · 5d

Navigating Academic Waters Webinar Series Continues!
Join us for, Expanding Funding Landscape: Non-Federal Funding Opportunities
OCT 8 / 1-2PM ET

Diversifying funding streams is becoming more important in hydrology and environmental science.
us06web.zoom.us/webinar/regi...

6 3

Reposted by Matthias Sprenger

Oliver Bossdorf @bossdorf.bsky.social · 5d

JOB ALERT! Do you also love large ecological experiments? Become coordinator of our big new geo-biosphere experiment in @terra-cluster.org. 7-year position! Follow link for details.
@gfoesoc.bsky.social
@britishecologicalsociety.org
@ecologicalsociety.bsky.social
@uslter.bsky.social

Please repost!

Stellenangebote: Universität Hohenheim

Aktuelle Stellenangebote der Universität Hohenheim.

www.uni-hohenheim.de

1 33 22

Reposted by Matthias Sprenger

Earth and Planetary Sciences at ETH Zurich @eth-eaps.bsky.social · 5d

We are empowering early career female researcher in Earth and Planetary Sciences: apply now for the ETH Zurich Judith McKenzie fellowships! Be part of a growing network of women shaping the future of science! Application deadline is 31 October 2025.

Funding opportunities

The Department of Earth and Planetary Sciences at ETH Zurich awards different fellowships for visiting post-tenure professors or research scientists, PhD students, and Postdocs.

eaps.ethz.ch

3 6

Matthias Sprenger @matthiasprenger.bsky.social · 7d

As I take with students our first water samples for isotope analyses in North Carolinian catchments I think of "a magic dwells in each beginning". It's fun to have a study site nearby and add to ongoing NCSU and USFS research.

11

Reposted by Matthias Sprenger

Science Vs @sciencevs.bsky.social · 10d

This moment from our latest episode with science writer @edyong209.bsky.social is 🔥

We asked Ed — how do we talk up the benefits of science in the face of government cuts? He told us that's the wrong approach. 🧪

Listen wherever, or watch on Spotify 👇

open.spotify.com/episode/7Evh...

8 340 650

Reposted by Matthias Sprenger

AGU Catchment Hydrology Technical Committee @agucatchhydro.bsky.social · 19d

🌊 This week in our #AGU25 #CatchmentHydrology Early Career series, we highlight: “Resilient Infrastructure & Disaster Response Center” by Abdul Mobin et al. (2025).

1 3 10

Reposted by Matthias Sprenger

AGU Hydrology H3S @agu-h3s.bsky.social · 9d

CATCHMENT CUP IS LIVE! Whose ready to learn about some new (and old) research catchments?!? First up: HEADWATER CATCHMENTS

VOTE HERE until 10/4: www.agu-h3s.org/rivercup

1 14 25

Matthias Sprenger @matthiasprenger.bsky.social · 10d

This comprehensive review was a massive collaborative effort initiated by the EU-COST Action WATSON. More info here: watson-cost.eu
Some coauthors on bluesky
@harshberia.bsky.social, @theoryofciaran.bsky.social, @danielepenna.bsky.social, @jlaknapp.bsky.social, @h2obabyts.bsky.social

Watson – Watson Cost

watson-cost.eu

2 3

Matthias Sprenger @matthiasprenger.bsky.social · 10d

Tracers also show that in many landscapes, trees use water from winter precipitation during the summer growing season, highlighting complex storage-and-release dynamics that simple water balance models might miss.

1

Matthias Sprenger @matthiasprenger.bsky.social · 10d

These models have led to important discoveries, like the "old water paradox"—stormflow is often dominated by older, stored water, not just the recent rainfall! This reveals the relevance of catchment storage.

Water mixing processes in surface waters and groundwaters (gw) are typically studied using tracer-aided mixing approaches. Typical end-members are illustrated in blue. Water compartments that are considered as end-members or mixtures, depending on the study, are shown in blue (end-member) and red (mixtures).

1

Matthias Sprenger @matthiasprenger.bsky.social · 10d

How do we track water from a raindrop to an aquifer? Tracer-aided mixing models use naturally occurring markers to identify water flow paths and quantify the contributions of different sources (like rain vs. groundwater) to streamflow.

Water exchange and mixing in the atmosphere-vegetation-soil continuum. Blue boxes represent end-members, red boxes represent mixtures, and the blue-red box represents both end-members and mixtures.

1

Matthias Sprenger @matthiasprenger.bsky.social · 10d

What is the Critical Zone (CZ)? 🌍 It's Earth's living skin—a dynamic interface where air, water, soil, plants, and rocks interact. Understanding water's journey through the CZ is vital for safeguarding our water resources and ecosystems.

Conceptual model of the water compartments (boxes) and water fluxes (italic text) within the Critical Zone that are typically studied using tracer-aided mixing models. Light blue boxes indicate compartments that are considered to be end-members. Red boxes indicate compartments that are considered as mixtures. Boxes with both colors represent compartments that are either end-members or mixtures. “Gw” stands for groundwater.

1