Travis Aerenson

About Me

Research

I study how clouds, moisture transport, and precipitation respond to a warming climate, how well Earth system models capture those processes, and what that means for water resources in the American West. My work combines satellite observations, reanalysis, and climate models, from simple analytical frameworks to large model ensembles and statistical emulation.

Where our precipitation comes from

Water vapor travels hundreds to thousands of miles between where it evaporates and where it falls, and theory and models predict that this journey lengthens as the climate warms. Applying the WAM2layersv3 moisture-tracking model to 35 years of ERA5 reanalysis, I built the Evaporation Dataset of Regional Precipitation Sources (E-DRoPS), which traces where precipitation falling in each of the 18 U.S. water resource regions (HUC2) originally evaporated. Using it, we found statistically significant trends across the Southwest (including the Upper and Lower Colorado, Rio Grande, and Arkansas-White-Red basins) and the southern Great Plains: water vapor travels farther and stays in the air longer before precipitating, and a larger share of it comes from oceans rather than land. In some regions vapor travels as much as 50 to 80 km farther and spends 2 to 4 more hours in the atmosphere than in the past. The trends follow shifts in the midlatitude storm track in the Upper Colorado basin and the Great Plains low-level jet in the Arkansas-White-Red basin. This is the first confirmation in the observational record of a change that had been predicted (Aerenson et al., 2026, GRL; news release). Ongoing work asks how moisture sources differ for extreme versus light precipitation and how global moisture transport constrains cloud feedback.

Snowpack and water supply in the mountain West

Winter snowpack in the Mountain West melts to supply rivers and reservoirs, so how it changes matters for agriculture, industry, and drinking water. Using ERA5-Land as a benchmark, we evaluated historical snow water equivalent (SWE) in Colorado, Utah, Wyoming, and Montana across CMIP6 models and three equally viable parameter tunings of NASA GISS-E3. ERA5-Land shows a much larger decline in March SWE between 1950–70 and 1995–2014 (about 20 mm) than the CMIP6 multimodel mean (about 7 mm), mostly because models underestimate how much warming reduces the share of fallen snow that stays on the ground. Splitting each SWE change into its precipitation, snow-fraction, and snow-retention components, and training a Gaussian process emulator on ERA5-Land to stand in for models’ land physics, I found that swapping in the emulator barely narrowed the spread among models. The disagreement comes instead from how models simulate regional temperature and precipitation minus evaporation, which in the Mountain West warmed 3.1 K in ERA5-Land but only about 1 K in the multimodel mean. The spread across the three GISS-E3 tunings alone was often comparable to the spread among different models, so parameter choices matter as much as model structure. CanESM5 and MRI-ESM2-0 were consistently best at reproducing historical SWE, though MRI-ESM2-0 appears to get the right answer for the wrong reasons. Selecting models by their historical snowpack climatology changed projections little, and selecting by historical trend broadened rather than narrowed the range of future outcomes (Aerenson et al., 2025, J. Climate). I am now working to connect the structure of winter storms over the ocean to the precipitation and snow they deliver after landfall.

Clouds in climate models

Timing. Even when a model gets the right cloud amount, it can misstate how much sunlight clouds reflect if it places them at the wrong time of day. We developed an analytical model that treats shortwave cloud radiative effect as the diurnal cycle of cloud liquid water path multiplied by the solar cycle, which splits a model’s bias into the daily mean, the amplitude of the daily cycle, its timing, and their interaction. Applied to the two CMIP6 models that saved sub-daily output (MPI-ESM1-2-HR and CanESM5) and compared with MAC-LWP satellite observations, the diurnal cycle produced biases in cloud radiative effect greater than 45% of the total bias, yet these were largely cancelled by compensating errors, so the models get roughly the right answer for the wrong reasons. A metric computed from observations alone, which we call phase susceptibility, shows where timing errors matter most: the stratocumulus decks off Peru, Namibia, California, and Australia, which also have some of the largest cloud radiative effect biases across 22 CMIP6 models. Tuning a 237-member E3SMv3 perturbed parameter ensemble to match observed mean liquid water path did not remove those biases. We argue that models should save the diurnal cycle of clouds, and note that the result bears on marine cloud brightening (Aerenson et al., 2025, Science Advances).

Feedbacks. Cloud feedbacks are the largest source of disagreement in climate sensitivity. Comparing nine CMIP5 and CMIP6 models with MISR satellite observations, whose stereo cloud-top heights separate thick high clouds from thin ones, I found that the strength of each of four cloud feedbacks (tropical high cloud, midlatitude low cloud, high-cloud altitude, and tropical low cloud) is statistically related to how the model simulates present-day clouds. Using MISR as an emergent constraint, I estimated a tropical high-cloud feedback of 0.00 ± 0.04 W/m²/K, much less negative than the −0.2 ± 0.2 of the most recent expert assessment, with a midlatitude low-cloud feedback of 0.18 ± 0.04, a high-cloud altitude feedback of 0.24 ± 0.02, and a tropical low-cloud feedback of 0.29 ± 0.09 W/m²/K. These constraints are one line of evidence rather than a replacement for others, but they suggest that future assessments should account for present-day cloud biases and that models could be tuned against satellite observations to reduce them (Aerenson & Marchand, 2025, JGR). Using paired experiments that raise or lower the solar constant by 4%, we compared temperature-mediated cloud feedbacks under solar forcing and CO₂ forcing, and found them broadly similar apart from a larger drop in low cloud under solar forcing, with asymmetric responses to warming and cooling (Aerenson & Marchand, 2024). In a companion paper we separated out the cloud adjustments that occur directly in response to the forcing, with a new method for the solar experiments. Increasing CO₂ reduces and lowers mid- and low-level clouds while solar forcing does not, solar forcing sharply reduces optically thin high clouds, and these adjustments still contribute significantly to the net cloud radiative effect after 150 years (Aerenson et al., 2024).

Detection. MISR shows a statistically significant cloud-top height trend only over the Southern Ocean so far; simulator output from two climate models suggests trends elsewhere should become detectable within a few years, and that a stable 20 to 30 year record is needed to separate them from natural variability (Aerenson et al., 2022, JGR). With collaborators I am also working to improve satellite estimates of cloud droplet number concentration.

What sets how hard it rains

Global mean precipitation is tightly constrained by the atmosphere’s energy budget, but the intensity with which it falls is not, and it remains one of the most uncertain aspects of model simulations. In work currently in review, we propose a mechanistic framework in which, because total precipitation is fixed, intensity is set by how precipitation is partitioned between the convection and large-scale cloud schemes, which in turn depends on the ratio of their precipitation efficiencies. Using the 251-member Nephele perturbed parameter ensemble of E3SMv3, we find that this efficiency ratio predicts precipitation intensity better than either efficiency alone, in most regions of the globe, and that the parameterized rate at which cloud droplets convert to rain (autoconversion) is the dominant control on both efficiencies. The framework offers a way to diagnose the persistent “too frequent, too light” precipitation bias in models, and I am working with ARM and satellite observations to constrain it.

Model uncertainty and emulation

Climate models can be tuned in many ways that all look reasonable. I collaborate on perturbed parameter ensembles, large sets of runs with systematically varied settings, to find which uncertain processes matter most. Examples include the Nephele ensemble for E3SMv3, which varies 25 aerosol-cloud parameters across 236 members and spans an aerosol-cloud effective radiative forcing of -2.6 to 0.7 W/m² (Nugent et al., 2026, JAMES), and a CAM6 ensemble used with observations to narrow the Southern Ocean cloud water response by 66 to 72% without ruling out high-sensitivity models (Werapitiya et al., 2025, J. Climate). Related collaborations evaluate aerosol-cloud interactions in a coupled reanalysis (Song et al., 2025, ACP) and energy transport in extreme-warming simulations (Poletti et al., 2024, PLOS Climate). I am a Co-I on the DOE EPSCoR project PROCEED.

Climate extremes under low-warming scenarios

As an undergraduate intern at NCAR, I computed ten standard extreme temperature and precipitation indices from three low-warming CESM1-CAM5 ensembles built around the Paris targets (1.5 °C, a 1.5 °C overshoot, and 2.0 °C). Significant changes emerged in nearly all indices; temperature extremes changed widely, while precipitation extremes were more affected by internal variability. Even scenarios separated by only half a degree of global warming had significantly different statistics of extremes (Aerenson et al., 2018, Environmental Research Letters).

Publications

Also see Google Scholar

[19] Werapitiya, G., Aerenson, T., McCoy, D., Brient, F., Elsaesser, G., Song, C., & Zelinka, M. How the hydrological cycle shapes global cloud feedback. (In prep for JGR: Atmospheres).

[18] Aerenson, T., Werapitiya, G., McCoy, D., Brient, F., Tsushima, Y., & Zelinka, M. Global moisture transport constrains cloud feedback in multiple large ensembles. (In prep for PNAS).

[17] Barahona, D., Aerenson, T., Gryspeerdt, E., & McCoy, D. Global vertically resolved reconstruction of cloud droplet number concentration combining satellite retrievals, reanalysis, and in situ observations. (In review Atmospheric Chemistry and Physics).

[16] Aerenson, T., McCoy, D., & Nugent, J. The efficiency ratio: convective and large-scale precipitation efficiency as a control on precipitation intensity. (In review Journal of Climate).

[15] Lou, J., McCoy, D., Caulton, D., Brown, H., Aerenson, T., & Nugent, J. M. Parameterized cloud and aerosol processes modulate the spatial distribution of extreme fire weather over North America. (In review JGR: Atmospheres).

[14] Aerenson, T., Barahona, D., & McCoy, D. Using a testbed model to identify sources of bias in spaceborne cloud droplet number concentration retrieval and propose new alternatives. (In review JGR: Atmospheres).

[13] Myers, T. A., Aerenson, T., McCoy, D., Zelinka, M. D., Nugent, J. M., & Brown, H. Cloud-Controlling Factors Reveal Mean-State Controls and Structural Biases in Climate Model Stratocumulus Physics. (Accepted, Geophysical Research Letters).

[12] Aerenson, T., McCoy, D., Keys, P., & Caulton, D. (2026). Precipitation Over the Contiguous United States Is Coming From Farther Away Than in the Past. Geophysical Research Letters, 53(14), e2026GL122565. https://doi.org/10.1029/2026GL122565
Press: The Guardian | Phys.org | Wyoming News | Cowboy State Daily interview | AGU newsletter (7/23/2026)

[11] Nugent, J. M., Brown, H., Kirby, A., McCoy, D. T., Allen, G., Aerenson, T., Burrows, S. M., Caulton, D., Fan, J., Feng, Y., Gettelman, A., Griswold, J., Jones, D. B., Leung, L. R., Mahfouz, N., Mikkelsen, A., Mülmenstädt, J., Qian, Y., Shan, Y., … Zhang, K. (2026). Overview of the Nephele Perturbed Parameter Ensemble for Aerosol-Cloud Interactions in E3SMv3. Journal of Advances in Modeling Earth Systems, 18(2), e2025MS004989. https://doi.org/10.1029/2025MS004989

[10] Song, C., McCoy, D., Molod, A., Aerenson, T., & Barahona, D. (2025). Signatures of aerosol-cloud interactions in GiOcean: a coupled global reanalysis with two-moment cloud microphysics. Atmospheric Chemistry and Physics, 25(21), 15567–15592. https://doi.org/10.5194/ACP-25-15567-2025

[9] Aerenson, T., McCoy, D., Elsaesser, G., Wu, J., Nugent, J. M., Brown, H., Mikkelsen, A., & Zelinka, M. D. (2025). Global models predict clouds at the wrong time of day: Does it matter for radiation and climate? Science Advances, 11(50), 3236. https://doi.org/10.1126/SCIADV.ADY3236
Press: University of Wyoming press release

[8] Werapitiya, G., McCoy, D. T., Elsaesser, G. S., Gettelman, A., Eidhammer, T., Aerenson, T., Song, C., & Wu, J. (2025). Extratropical Cloud Feedback Constrained by Cloud Sources and Sinks in Cyclones. Journal of Climate, 1(aop). https://doi.org/10.1175/JCLI-D-24-0607.1

[7] Aerenson, T., McCoy, D., Elsaesser, G., & Wu, J. (2025). Drivers of Model Spread in Snowpack Changes Across the American Mountain West. Journal of Climate, 38(22), 6605-6619. https://doi.org/10.1175/JCLI-D-25-0103.1

[6] Aerenson, T. & Marchand, R. (2025). How Do Differences in the Simulation of Present-Day Clouds Affect Cloud Feedbacks? Journal of Geophysical Research: Atmospheres, 130(12), e2025JD044020. https://doi.org/10.1029/2025JD044020

[5] Aerenson, T. & Marchand, R. (2024). Cloud Responses to Abrupt Solar and CO2 Forcing: 1. Temperature Mediated Cloud Feedbacks. Journal of Geophysical Research: Atmospheres, 129(12), e2023JD040296. https://doi.org/10.1029/2023JD040296

[4] Aerenson, T., Marchand R., & Zhou, C. (2024). Cloud Responses to Abrupt Solar and CO2 Forcing: 2. Adjustment to Forcing in Coupled Models. Journal of Geophysical Research: Atmospheres, 129(12), e2023JD040297. https://doi.org/10.1029/2023JD040297

[3] Poletti, A. N., W Frierson, D. M., Aerenson, T., Nikumbh, A., Carroll, R., Henshaw, W., & Scheff, J. (2024). Atmosphere and ocean energy transport in extreme warming scenarios. PLOS Climate, 3(2), e0000343. https://doi.org/10.1371/JOURNAL.PCLM.0000343

[2] Aerenson, T., Marchand, R., Chepfer, H., & Medeiros, B. (2022). When Will MISR Detect Rising High Clouds? Journal of Geophysical Research: Atmospheres, 127(2), e2021JD035865. https://doi.org/10.1029/2021JD035865
Press: Skeptical Science

[1] Aerenson, T., Tebaldi, C., Sanderson, B., & Lamarque, J.F. (2018). Changes in a suite of indicators of extreme temperature and precipitation under 1.5 and 2 degrees warming. Environmental Research Letters ERL 13 035009 https://doi.org/10.1088/1748-9326/aaafd6

Presentations

[28] Aerenson, T., McCoy, D., Caulton, D., Myers, T., Nugent, J., Lou, J., Ross, T. 2026: “The impact of cloud processes on predictability across the Earth system” E3SM all-hands meeting plenary session, Bethesda, MD. Oral

[27] Aerenson, T. 2026: “Moisture Transport, Precipitation, and Snowpack: what global models tell us about the water cycle and how it is changing” Florida State University Department of Earth, Ocean, and Atmospheric Science, Tallahassee FL. Invited Seminar

[26] Aerenson, T. 2025: “Remote causes of extreme precipitation in the Western US” AGU Fall Meeting, New Orleans, LA. Poster

[25] Aerenson, T., Barahona, D., McCoy, D. 2025: “Cloud adiabaticity and droplet size distribution cause bias in spaceborne cloud droplet number concentration retrieval.” AGU Fall Meeting, New Orleans, LA. Poster

[24] Aerenson, T. 2025: "Snowpack and atmospheric moisture transport: how the water cycle is changing in a warming world." University of Wyoming Department of Atmospheric Science Seminar, Laramie, WY. Seminar

[23] Aerenson, T., D. McCoy, G. Elsaesser, J. Wu, J. Nugent, H. Brown, M. Zelinka, S. Burrows 2025: "Are we simulating clouds at the wrong time of day?" Gordon Research Conference on Radiation and Climate, Lewiston, ME. Poster

[22] Aerenson, T., D. McCoy, G. Elsaesser, J. Wu, J. Nugent, H. Brown, M. Zelinka, S. Burrows 2025: "Are we simulating clouds at the wrong time of day?" Gordon Research Seminar on Radiation and Climate, Lewiston, ME. Oral

[21] Aerenson, T., R. Marchand, T. Ackerman 2024: “MISR Observed Trends in Cloud Top Height and Constraints on High Cloud Altitude Feedbacks” AGU Fall Meeting: Terra: 25 Years of the Earth Observing System Flagship Observatory, Washington, DC. Poster

[20] Aerenson, T., D. McCoy, G. Elsaesser 2024: “Causes of Model Spread in Predictions of Hydroclimate in the Mountain West” AGU Fall Meeting: Hydroclimate and Extremes in the Western United States in a Changing Climate, Washington, DC. Poster

[19] Aerenson, T., D. McCoy 2024: "How Much Does the Cloud Diurnal Cycle Impact SWCRE?" Micro2Macro Origins of Climate Change Uncertainty Workshop, Laramie, WY. Poster

[18] Aerenson, T., D. McCoy, G. Elsaesser 2024: "How Might we Improve Predictions of Regional Hydroclimate?" Oxford Workshop on Model Uncertainty, Oxford, UK. Oral

[17] Aerenson, T., D. McCoy, G. Elsaesser 2024: “Can We Do Better at Predicting Regional Hydroclimate?” CESM Workshop, Boulder, CO. Oral

[16] Aerenson, T. 2023: "The Relationship Between Simulated Present-Day Cloud Attributes and Cloud Feedbacks" University of Washington Department of Atmospheric Sciences Colloquium, Seattle, WA. PhD Defense

[15] Aerenson, T. R. Marchand, C. Zhou 2023: "Cloud Adjustments to Solar and CO2 Forcing in Coupled Models" CFMIP Meeting, Paris, FR. Poster

[14] Aerenson, T. R. Marchand 2023: "The Contribution of Mean-State Bias to Cloud Feedbacks in Climate Models" CFMIP Meeting, Paris, FR. Poster

[13] Aerenson, T. R. Marchand 2023: "The Contribution of Mean-State Bias to Cloud Feedbacks in Climate Models" University of Wyoming Department of Atmospheric Science Seminar, Laramie, WY. Invited Seminar

[12] Aerenson, T. R. Marchand 2023: "Using ISCCP and MISR Satellite Simulators to Understand Cloud Feedbacks" NASA GSFC CPC Seminar, Greenbelt, MD. Virtual Invited Seminar

[11] Aerenson, T., R. Marchand, C. Zhou 2022: “Cloud Response to Abrupt Changes in Solar Forcing and CO2 Concentrations” AGU Fall Meeting: Advances in Solar Radiation Modification Research, Chicago, IL. Poster

[10] Aerenson, T., R. Marchand, C. Zhou 2022: “Cloud Response to Abrupt Changes in Solar Forcing and CO2 Concentration” University of Washington Department of Atmospheric Sciences Seminar on Atmospheric Physics and Chemistry, Seattle, WA. Seminar

[9] Aerenson, T., R. Marchand 2022: “Cloud Response to Abrupt Changes in Solar Forcing and CO2 Concentrations” CFMIP Meeting: Cloud Processes and Radiative Feedbacks, Seattle, WA. Oral

[8] Aerenson, T. 2021: “Cloud Rapid Adjustments and Feedbacks to Abrupt Changes in Solar and CO2 Induced Forcings” AGU Fall Meeting: Advances in Climate Engineering Research. Virtual Poster

[7] Aerenson, T., R. Marchand 2021: “Cloud Rapid Adjustments and Feedbacks to Abrupt Changes in Solar and CO2 Induced Forcing” CFMIP Meeting. Virtual Poster

[6] Aerenson, T. 2021: “When Will MISR Detect Rising High Clouds?” University of Washington Department of Atmospheric Sciences Physics and Chemistry Seminar. Virtual Seminar

[5] Aerenson, T., R. Marchand 2021: “Time of Emergence: When Will We See High Clouds Get Higher?” AMS Annual Meeting. Remote oral presentation and discussion session

[4] Aerenson, T., R. Marchand, 2020: “Time of Emergence: When do Climate Models Predict Rising Cloud-Top-Height (CTH) Should be Detected by MISR?” CFMIP Meeting on Clouds, Precipitation, and Climate Sensitivity. Remote submitted slide and discussion session

[3] Aerenson, T., R. Marchand, 2020: “When will we see high clouds get higher?” MISR Science Team Meeting, Pasadena, CA. Oral

[2] Aerenson, T., 2019: “Climate Models and Climate Change Reversibility” Colorado College Physics Department Senior Seminar Series, Colorado Springs, CO. Seminar

[1] Aerenson, T., C. Tebaldi, B. Sanderson, J.F. Lamarque, 2017: “Climate Extremes in Low Warming Scenarios” NCAR CGD Integrated Assessment Modelling Weekly Meeting, Boulder, CO. Oral

Teaching & Mentoring

Teaching

• Guest Lecturer, University of Wyoming, Oct 2024-Present: ATSC 2100 Global Warming: The Science of Humankind's Energy Consumption Impacting Climate; ATSC 5040 Global Physical Climatology
• Teaching Assistant, University of Washington, 2021: ATM S 100 Climate Justice & Energy Solutions
• Math and Physics Tutor, Quantitative Reasoning Center, Colorado College, 2017-2019

Mentoring

• Tobias Ross: postdoc supervisor
• Timothy Considine: undergraduate research advisor (graduated 2026)
• Geethma Werapitiya: mentor (PhD, 2025)
• Kyle Johnson: mentor and PhD committee member
• Carson Turner: mentor and MS committee member
• Kinsale Day: MS committee member

Service and community

• Postdoc Representative (2024-2025) and Strategic Planning Committee (2025), University of Wyoming Department of Atmospheric Science
• Peer-to-peer Mentoring Coordinator (2022-2023) and Diversity and Inclusion Group Coordinator (2021-2023), University of Washington Department of Atmospheric Sciences
• Panelist: UW College of the Environment Pathways to Postdoc (2025); UWyo College of Engineering and Physical Sciences Graduate Program Panel (2024)
• Certificate of Distinguished Service (2022), University of Washington Department of Atmospheric Sciences
• Reviewer for JGR: Atmospheres, GRL, JAMES, ACP, Nature Communications, Science Advances, and others; proposal reviewer for the NASA Postdoctoral Program