Western US Research Articles

Exploring Climate-Driven agricultural water shortages in a Snow-Fed basin using a water allocation model and Machine learning

Water allocation models (WAM) can capture complex feedbacks between water infrastructure, water supply, water demand, and water rights structure. They are an important tool for water managers to reduce conflict and aid in future water resource planning. Future streamflow reductions and shifts in timing are anticipated in snow-dominated watersheds of the western United States due to projected reductions in snow water storage. These basins tend to be water-limited and rely on groundwater to support agricultural demand. Water managers in the snow-fed, agricultural Walker River Basin (WRB) in California and Nevada use a dynamically coupled WAM and physically based groundwater flow model as a decision support tool. Large computation time is required to simulate groundwater dynamics, limiting the number of model scenarios that can feasibly be run to test alternative management scenarios under climate change. Machine learning (ML) techniques have been used to capture complex nonlinear hydrologic dynamics in other contexts, but to our knowledge there have not been efforts to use ML to increase computational efficiency in modeling groundwater responses to agricultural water use. We use extreme gradient boosting to replace the groundwater sub-model for the WRB decision support tool, resulting in accurate estimates of reservoir storage (monthly error almost always < 5% of reservoir capacity), streamflow (total outflow errors from −3% to 20% depending on scenario) and agricultural water shortages (-12% to 18% error in total shortage depending on scenario). For comparison, we found that omitting groundwater modeling from the WAM results in a larger overprediction of outflows (2 to 58%) and underprediction in agricultural water shortages (0 to 41%) across a range of scenarios. We use this validated ML method to explore the sensitivity of agricultural water availability to a range of streamflow volumes and streamflow timings. Results indicate that agricultural water shortages are more sensitive to annual streamflow volume in comparison to shifts in streamflow timing. Our model shows that reservoirs can buffer the impact of altered flow timing on agricultural production, except under extreme conditions in which all precipitation falls as rain (no snow storage) and streamflow peaks in the winter. Future work will explore these thresholds in more detail. While the results here are site-specific, the methodology for coupling a WAM with a ML representation of surface–groundwater interactions can be applied to a wide variety of water resource modeling applications. These results also provide a stark example of the importance of snowpack storage and groundwater management to western US water infrastructure.

The Increasing Role of Seasonal Rainfall in Western U.S. Summer Streamflow

AbstractSummer streamflow variations strongly affect water supply reliability and ecological functioning of western U.S. (WUS) streams. Traditional snow‐based forecasts of summer streamflow are becoming less accurate with warming‐induced reductions in winter snow accumulation. This reflects a rising importance of competing runoff‐generating processes in controlling summer streamflow variations, primarily an increasing role of rainfall in contrast to snowmelt. Here, based on a snowmelt‐rainfall tracking algorithm applied to two hydrological models, we show that cool‐season rainfall provides an important volumetric contribution to summer streamflow for many WUS streams in the current climate, and this contribution will increase under climate warming, especially in years with warm snow droughts and abnormally dry summers. We also show that seasonal rainfall (warm‐/cool‐seasons) dominates the variability of summer streamflow across ∼70% area of WUS. We show that an increasing warm‐season rainfall contribution to summer streamflow (largely replacing snowmelt) results in reduced summer streamflow predictability.

Mountains in Oman Can Store Huge Amounts of CO2 if a Way Can Be Found Into the Tight Rock

Every year, the unusual mix of minerals in the Hajar Mountains near the coast of Oman traps 100,000 tons of carbon in the rock. That estimate is a tiny fraction of the potential of the mountain range and a few others like it in the world. Based on decades of work by geologists studying this unique formation—known as the Samail Ophiolite—the highly reactive rocks called peridotites can theoretically trap one-half ton of CO2 per ton of that rock in the Hajar Mountains, which extend into the UAE. A leading voice among the researchers who made the 100,000-ton estimate and developed ideas for removing trillions of tons of carbon from the atmosphere is Peter Kelemen, a professor at Columbia University in New York City. In a 2019 story in Scientific American, Kelemen said that if it is possible to speed the pace of mineralization “by a factor of a million”—something he thinks is doable with a bit of engineering—“then you end up with a billion tons of CO2 per cubic kilometer of rock per year.” That potential inspired an Omani entrepreneur, Talal Hasan, to start a company based on the work of Kelemen and colleagues including Juerg Matter, a geochemist now working at the University of Southampton in England. The result was startup 44.01, named for the molecular mass of carbon, where Kelemen serves as an advisor and Matter works part time. The company’s website describes its plan: “Carbon mineralization in peridotite is happening all the time—we simply speed up the natural process.” It is a simple-sounding goal. But what it will take to realize the vast potential is anything but simple. In an interview, Kelemen identified why: “The main concern is that the rocks are not very porous.” Or very permeable, which helps explain why large amounts of highly reactive elements have remained untouched over the 96 million years these reactive minerals have been on land. The hard rocks in this formation, particularly magnesium-rich olivine, were produced when magma flowed up from the mantle to a mid-ocean ridge where it cooled and the thick layer of rock spread out toward what is now Oman and the UAE, and something extraordinary happened. When it reached the boundary between two plates, it normally would have returned to the depths of the earth. Instead, it collided with another ophioplite and was ultimately thrust up onto land. When the theory of plate tectonics transformed geological thinking during the second half of the 20th century, this giant anomaly became a destination for geologists eager to study a rare outcrop of rock from the mantle, including Kelemen. They also could have gone to New Guinea to look at a similar display of mantle rock or scattered sites in the western US, among other places, where ophiolites are present. But the Hajar Mountains comprise the biggest ophiolite of them all. Speeding carbon mineralization from geological rates to the pace needed to help stop global warming poses a fundamental problem common to unconventional oil and gas—creating flow paths in tight rock on a limited budget.

AP-452680-2 GEOSPATIAL ANALYSIS OF BYSTANDER-INITIATED AUTOMATED EXTERNAL DEFIBRILLATOR UTILIZATION IN A LARGE US COMMUNITY

Bystander use of automated external defibrillators (AED) can improve survival from out-of-hospital sudden cardiac arrest (SCA), but utilization can be variable. We hypothesized that AED distribution and utilization is determined by race/ethnicity and socioeconomic status of neighborhoods. To evaluate the distribution and utilization of bystander-initiated AED use for SCA events in a single large US community. All consecutive SCA cases with resuscitation attempted by emergency medical services (EMS) in Ventura County, CA (pop. 848,112, 43% Hispanic) were prospectively identified during 2015-2022. AED locations were obtained from the Ventura County Healthcare Agency Registry. Census tracts in the county (US Census 2018 estimates), were dichotomized as above or below the median census tract household income categorized as SES1 (income <median income) and SES2 (income >median income). AED and SCA event locations were geocoded and merged with the 2018 census tract geospatial map. The Euclidean distance (meters) between these two locations was calculated. AEDs were considered in viable proximity if located within 300 meters of the SCA event. Differences were analyzed using univariate statistical methods. Among 2515 cases (70.5±16.33 y, 65% male) of whom 684 (27%) were Hispanic, 782 (31%) were in viable proximity of an AED (Figure). There were no ethnicity-based differences in viable AED proximity (Hispanic n= 225(33%), non-Hispanic n=557 (30%), p=0.23). A total of 1442 (57%) cases (70±16.17 y, 65% male) were categorized as SES1 and 1070 (43%) cases (71±16.52 y, 65% male) as SES2. No significant differences in viable AED proximity were identified between SES1 (n= 434, 30%), and SES2 (n=348(33%); p=0.19). A total of 1361(54%) received bystander CPR, but bystander-initiated AED use was observed in 75 (3%). AED use was observed in 75/782 (9%) SCA cases in viable proximity to an AED of whom 3.1% were Hispanic and 12.2% non-Hispanic (p <0.001). SES2 cases had higher AED utilization than SES1 cases (12.4% vs. 7.8%, p = 0.02). In this western US community, AEDs in viable proximity to SCA events were distributed equitably, and not affected by neighborhood SES or ethnicity. However, bystander-initiated use for AEDs in close proximity was under 10%, indicating significant missed opportunities. These findings highlight the critical need to enhance community-wide education in bystander-initiated CPR and AED use for further improvement in SCA survival outcomes.

Comparing Adjoint Waveform Tomography Models of California Using Different Starting Models

AbstractAdjoint waveform tomography (AWT) sits at the cutting edge of seismic tomography on local, regional, and global scales. However, the choice in starting model may have a significant impact on the final inversion results. In this paper, we present 3 AWT models of California that are based on different starting models. We chose three models that were inverted at different scales: SPiRaL, a global travel‐time tomography model (Simmons et al., 2021, 10.1093/gji/ggab277), CSEM_NA, a regional adjoint tomography model of North America and the North Atlantic (Krischer et al., 2018, 10.1029/2017JB015289), and WUS256, a regional adjoint tomography model of the western US (Rodgers et al., 2022, https://doi.org/10.1029/2022JB024549). We then inverted three AWT models using the same source and receiver set. We ran each model over three period bands: 30–100 s, 25–100 s, and 20–80 s. Once the iterations were finalized, we used five methods of testing model similarity in both the model and data space. We conclude that the choice of starting model has a minimal impact on long wavelength models if an appropriate multi‐scale inversion approach is used.

Exploring the Factors Controlling the Long‐Term Trend (1988–2019) of Surface Organic Aerosols in the Continental United States by Simulations

AbstractObserved surface organic aerosols (OA) concentrations slightly increased in the western US (WUS) but significantly decreased in the eastern US (EUS) in summer, and continuously decreased in winter over the US region. To understand the driving factors for the long‐term surface OA trend, we apply a revised version of the Community Atmosphere Model version 6 with comprehensive tropospheric and stratospheric chemistry representation, considering the heterogeneous formation of isoprene‐epoxydiol‐derived secondary organic aerosols (SOAIE) and fast photolysis rate of monoterpene‐derived secondary organic aerosols (MTSOA) to diagnose the OA evolution in 1988–2019. Compared to older versions, the revised model better reproduces the climatology, seasonal cycle, and long‐term trend of surface OA as evaluated against the Interagency Monitoring of Protected Visual Environments measurements. We find the decrease in EUS summertime OA is likely attributed to the interplay between SOAIE and MTSOA. With anthropogenic emissions reduction, primary organic aerosols (POA) declined, SOAIE decreased along with sulfate, while MTSOA increased along with biogenic emissions driven by a warming climate. POA from wildfires with a significant trend of 2.9% yr−1 and considerable interannual variation of 62.8% drive the statistically insignificant but increasing WUS summertime OA, while anthropogenic POA dominates the decreasing wintertime OA in the US. Through sensitivity experiments, we find MTSOA show linear responses to the increasing monoterpenes emissions and negligible responses to NOx emissions reduction due to the mutual offsets between MTSOA components from different oxidation pathways. This study reveals the increasingly important role of MTSOA in summertime OA under a warming climate.

A multi-satellite framework to rapidly evaluate extreme biosphere cascades: The Western US 2021 drought and heatwave.

The increasing frequency and intensity of climate extremes and complex ecosystem responses motivate the need for integrated observational studies at low latency to determine biosphere responses and carbon-climate feedbacks. Here, we develop a satellite-based rapid attribution workflow and demonstrate its use at a 1-2-month latency to attribute drivers of the carbon cycle feedbacks during the 2020-2021 Western US drought and heatwave. In the first half of 2021, concurrent negative photosynthesis anomalies and large positive column CO2 anomalies were detected with satellites. Using a simple atmospheric mass balance approach, we estimate a surface carbon efflux anomaly of 132 TgC in June 2021, a magnitude corroborated independently with a dynamic global vegetation model. Integrated satellite observations of hydrologic processes, representing the soil-plant-atmosphere continuum (SPAC), show that these surface carbon flux anomalies are largely due to substantial reductions in photosynthesis because of a spatially widespread moisture-deficit propagation through the SPAC between 2020 and 2021. A causal model indicates deep soil moisture stores partially drove photosynthesis, maintaining its values in 2020 and driving its declines throughout 2021. The causal model also suggests legacy effects may have amplified photosynthesis deficits in 2021 beyond the direct effects of environmental forcing. The integrated, observation framework presented here provides a valuable first assessment of a biosphere extreme response and an independent testbed for improving drought propagation and mechanisms in models. The rapid identification of extreme carbon anomalies and hotspots can also aid mitigation and adaptation decisions.

ABORTION AND NEONATAL MORTALITY DUE TO TOXOPLASMA GONDII IN BIGHORN SHEEP (OVIS CANADENSIS).

Low lamb recruitment can be an obstacle to bighorn sheep (Ovis canadensis) conservation and restoration. Causes of abortion and neonate loss in bighorn sheep, which may affect recruitment, are poorly understood. Toxoplasma gondii is a major cause of abortion and stillbirth in domestic small ruminants worldwide, but no reports exist documenting abortion or neonatal death in bighorn sheep attributable to toxoplasmosis. Between March 2019 and May 2021, eight fetal and neonatal bighorn lamb cadavers from four western US states (Idaho, Montana, Nebraska, and Washington) were submitted to the Washington Animal Disease Diagnostic Laboratory for postmortem examination, histologic examination, and ancillary testing to determine the cause of abortion or neonatal death. Necrotizing encephalitis characteristic of toxoplasmosis was identified histologically in six of eight cases, and T. gondii infection was confirmed by PCR in five cases with characteristic lesions. Other lesions attributable to toxoplasmosis were pneumonia (3/5 cases) and myocarditis (2/5 cases). Protozoal cysts were identified histologically within brain, lung, heart, skeletal muscle, adipose tissue, or a combination of samples in all five sheep with PCR-confirmed T. gondii infections. Seroprevalence of T. gondii ranged from 40-81% of adult females sampled in the Washington population in October and November 2018-2021, confirming high rates of exposure before detection of Toxoplasma abortions in this study. Of 1,149 bighorn sheep postmortem samples submitted to Washington Animal Disease Diagnostic Laboratory between January 2000 and May 2021, 21 of which were from fetuses or neonates, a single case of chronic toxoplasmosis was diagnosed in one adult ewe. Recent identification of Toxoplasma abortions in bighorn sheep suggests that toxoplasmosis is an underappreciated cause of reproductive loss. Abortions and neonatal mortalities should be investigated through postmortem and histologic examination, particularly in herds that are chronically small, demographically stagnant, or exhibit reproductive rates lower than expected.

Disease Surveillance of Cougars (Puma concolor) in Utah, USA.

Adult, free-ranging cougars (Puma concolor) were sampled in three regions of Utah, US, from 2018 to 2021. A total of 68% (23/34) of the sampled cougars had antibodies to feline parvovirus, 15% (5/33) to canine distemper virus, 18% (6/34) to calicivirus, and 22% (8/37) to Yersinia pestis. Forty-one percent (13/32) had IgG antibodies to Toxoplasma gondii and 6% (2/33) to feline immunodeficiency virus, and 3% (1/32) were positive for Dirofilaria immitis (heartworm) antigen. All were seronegative for Toxoplasma gondii IgM, feline enteric coronavirus, SARS-CoV-2, feline leukemia virus, feline herpesvirus, and Francisella tularensis. Tapeworms and Toxascaris leonina eggs were detected in the feces. The disease exposures detected were similar to what has been reported from cougar populations in other western US states, and the current level of exposures is unlikely to have a negative impact on the state's population.

Climate change, tree demography, and thermophilization in western US forests

Climate change is driving widespread changes in ecological communities. Warming temperatures often shift community composition toward more heat-tolerant taxa. The factors influencing the rate of this "thermophilization" process remain unclear. Using 10-y census data from an extensive forest plot network, we show that mature tree communities of the western United States have undergone thermophilization. The mean magnitude of climate warming over the 10-y study interval was 0.32 °C, whereas the mean magnitude of thermophilization was 0.039 °C. Differential tree mortality was the strongest demographic driver of thermophilization, rather than growth or recruitment. Thermophilization rates are associated with recent changes in temperature and hydrologic variables, as well as topography and disturbance, with insect damage showing the strongest standardized effect on thermophilization rates. On average, thermophilization occurred more rapidly on cool, north-facing hillslopes. Our results demonstrate that warming temperatures are outpacing the composition of western US forest tree communities, and that climate change may erode biodiversity patterns structured by topographic variation.

Uncertainty quantification of machine learning models to improve streamflow prediction under changing climate and environmental conditions

Machine learning (ML) models, and Long Short-Term Memory (LSTM) networks in particular, have demonstrated remarkable performance in streamflow prediction and are increasingly being used by the hydrological research community. However, most of these applications do not include uncertainty quantification (UQ). ML models are data driven and can suffer from large extrapolation errors when applied to changing climate/environmental conditions. UQ is required to quantify the influence of data noises on model predictions and avoid overconfident projections in extrapolation. In this work, we integrate a novel UQ method, called PI3NN, with LSTM networks for streamflow prediction. PI3NN calculates Prediction Intervals by training 3 Neural Networks. It can precisely quantify the predictive uncertainty caused by the data noise and identify out-of-distribution (OOD) data in a non-stationary condition to avoid overconfident predictions. We apply the PI3NN-LSTM method in the snow-dominant East River Watershed in the western US and in the rain-driven Walker Branch Watershed in the southeastern US. Results indicate that for the prediction data which have similar features as the training data, PI3NN precisely quantifies the predictive uncertainty with the desired confidence level; and for the OOD data where the LSTM network fails to make accurate predictions, PI3NN produces a reasonably large uncertainty indicating that the results are not trustworthy and should avoid overconfidence. PI3NN is computationally efficient, robust in performance, and generalizable to various network structures and data with no distributional assumptions. It can be broadly applied in ML-based hydrological simulations for credible prediction.

Alpine hillslope failure in the western US: Insights from Chaos Canyon landslide, Rocky Mountain National Park USA

Alpine hillslope failure in the western US: Insights from Chaos Canyon landslide, Rocky Mountain National Park USA

Comparison of Soil Nitrate and Phosphorus Concentrations Prior to and Five to Eleven Years after Recycled Water Irrigation

Increasing demand on water supplies in western US and more stringent wastewater discharge standards have made recycled water a common water source for irrigating urban green spaces. Studies are needed to evaluate nitrate-N leaching potential and phosphorus (P) movement along soil profile when recycled wastewater is used for irrigation. We collected and analyzed soil samples at the commencement and 5 and 11 years after recycled water irrigation on 2 golf courses, 5 metropolitan parks, and 1 school ground. Samples were taken at 0-20, 20-40, 40-60, 60-80, and 80-100 cm depths on golf courses and at 0-20 and 20-40 cm depths at other locations. Soil samples were tested for soil pH, soil nitrate, and AB-DTPA extracted P level. No increase in soil nitrate-N was observed over 5 and 11 years with recycled water irrigation, suggesting leaching of nitrogen to the groundwater was not a great concern. Soil P concentration at the surface soil depth was the highest 11 years after recycled water irrigation. Moreover, soil P increasing below the surface layer was observed on sites with sandy soil, suggesting long-term recycled water irrigation could impose some risk for P leaching on sandy soil.

Human–river relationships depend on human–human relationships: River and watershed organizations in three western US states

AbstractHuman–nature relationship concepts are held collectively within society and guide environmentally oriented actions. This article explores the roles played by environmental organizations, particularly those focused on rivers and watersheds, in catalyzing interaction and action driven by human–river relationship goals. Interviews were conducted with representatives from 64 river and watershed organizations in Montana, Utah, and Wyoming in 2022. Organizational representatives were asked about mission focus areas, human–river relationships, the knowledge they draw upon to guide their efforts, and factors and obstacles that enable and constrain their progress and success. These qualitative data reveal a strong orientation toward steward and partner types of human–nature relationship concepts; however, there are discrepancies in conceptual interpretations. For river and watershed organizations in the US Intermountain West, human–river relationship goals depend strongly on human–human relationships in the form of diverse knowledge integration, collaboration, partnerships, trust, and communication in order to achieve their river‐related goals.

Ensemble species distribution modeling of Culex tarsalis (Diptera: Culicidae) in the continental United States.

West Nile virus (WNV) is the primary mosquito-borne disease in the United States and has had case reports every year since its introduction in 1999. As such, it is critical that we characterize the distribution of WNV vectors. Estimates of Culex tarsalis Coquillett species distribution, a major WNV vector, are scarce. We used ensemble distribution modeling to estimate habitat suitability for this species across the contiguous United States by consolidating presence data from four publicly available mosquito trapping data servers. The central plains region and much of the western US were estimated to have high habitat suitability. We identified multiple metrics of temperature and precipitation to be important in predicting the occurrence of Cx. tarsalis in a given geographic area. Furthermore, we observed habitat suitability for Cx. tarsalis to be significantly higher in areas with a high incidence of West Nile neuroinvasive disease compared to areas with low WN disease incidence, suggesting that Cx. tarsalis is present in regions with a high incidence of disease.

The magnitude and pace of photosynthetic recovery after wildfire in California ecosystems

Wildfire modifies the short- and long-term exchange of carbon between terrestrial ecosystems and the atmosphere, with impacts on ecosystem services such as carbon uptake. Dry western US forests historically experienced low-intensity, frequent fires, with patches across the landscape occupying different points in the fire-recovery trajectory. Contemporary perturbations, such as recent severe fires in California, could shift the historic stand-age distribution and impact the legacy of carbon uptake on the landscape. Here, we combine flux measurements of gross primary production (GPP) and chronosequence analysis using satellite remote sensing to investigate how the last century of fires in California impacted the dynamics of ecosystem carbon uptake on the fire-affected landscape. A GPP recovery trajectory curve of more than five thousand fires in forest ecosystems since 1919 indicated that fire reduced GPP by [Formula: see text] g C m[Formula: see text] y[Formula: see text]([Formula: see text]) in the first year after fire, with average recovery to prefire conditions after [Formula: see text] y. The largest fires in forested ecosystems reduced GPP by [Formula: see text] g C m[Formula: see text] y[Formula: see text] (n = 401) and took more than two decades to recover. Recent increases in fire severity and recovery time have led to nearly [Formula: see text] MMT CO[Formula: see text] (3-y rolling mean) in cumulative forgone carbon uptake due to the legacy of fires on the landscape, complicating the challenge of maintaining California's natural and working lands as a net carbon sink. Understanding these changes is paramount to weighing the costs and benefits associated with fuels management and ecosystem management for climate change mitigation.

Characterizing the development and drivers of 2021 Western US drought

Understanding the drivers of the major 2021 drought conditions across the Western U.S. (WUS) is important to develop future resilience and adaptation measures. While evapotranspiration (ET) tends to increase in a warming climate when water is available, areas with low precipitation would evaporate less water than expected, as seen in the 2021 drought. This study examines the relative roles of higher temperatures and lower precipitation, as well as anthropogenic forcing (e.g. increased greenhouse gases and land-use land-cover change) to the 2021 drought across the WUS. Using observations, the anomalously dry 2021 soil moisture is mainly tied to precipitation deficit, rather than higher temperatures, suggesting that an increase in ET in response to higher temperatures (i.e. saturation deficit) depends on water availability. Therefore, high temperatures may play only a secondary role in driving the 2021 drought condition across the WUS. Additionally, a suite of variable infiltration capacity model experiments confirms that the reduced precipitation in 2021 has caused negative soil moisture anomalies. Based on the Coupled Model Intercomparison Project Phase 6 experiments, anthropogenic forcing dramatically increases the risk of the extreme 2021 dryness, with risk ratio being 73.91, 12.78 and 25.81 for temperature, precipitation, and soil moisture respectively. Therefore, the extreme drought is not explained by natural forcing (e.g. solar irradiance and volcanic eruption) alone. Rather, anthropogenic forcing (e.g. increased greenhouse gases and land-use land-cover change) has increased the risk of this drought condition by approximately 26 times in terms of soil moisture compared with a world without this forcing.

Market-oriented solutions for groundwater commons through collective-action

Groundwater scarcity poses threats to communities across the globe, and effectively managing those challenges requires designing policy that achieves institutional fit. Collective action is well-suited to match rules with local context, and multiple pathways exist for communities to achieve reductions in groundwater use. To better understand how local conditions influence rule design, we examine two groundwater-reliant communities in the Western US that engaged in collective-action to arrive at distinct groundwater management rules. We consider: what drove stakeholders in Northwestern Kansas (NWKS) and San Luis Valley, Colorado (SLV) to adopt local groundwater policies, and why were different management pathways chosen? Further, why is more heterogeneity observed between local management organizations in SLV as compared to NWKS? To investigate these questions, we employ grounded theory to interpret the rules in reference to local hydro-agro-economic statistics and interviews with stakeholders (n = 19) in each region selected by expert sampling. We find that the additional goals of groundwater rules in SLV, partially driven by key contrasts in the resource system compared to NWKS, and higher resource productivity in SLV, creates both the need for and efficacy of a price-centered policy. Furthermore, variation in the resource systems and associated farm characteristics between subdistricts drives higher heterogeneity in rule design between local management districts in SLV compared to NWKS. More generally, we find the local flexibility afforded through the collective-action process as critical, even if it were to arrive at alternative, non-economic based incentives.

Changing Seasonality of Annual Maximum Floods over the Conterminous US: Potential Drivers and Regional Synthesis

Understanding the flood-generating mechanisms that influence flood seasonality in a region provides information on setting up relevant contingency measures. Although former studies estimated flood seasonality at regional/continental scale, limited/no studies have investigated the climate/basin drivers that influence the changes in flood seasonality. Considering this, the current study performed two analyses: (1) estimated the changes in the seasonality of annual maximum floods (AMF) between pre- and post-1970 across Hydroclimate Data Network basins over the conterminous US, and (2) identified the predictors that influence the change in the seasonality from a set of climate and geomorphic variables. Significant changes in the AMF seasonality were noted for approximately half of the basins in the eastern US, but low to no change was found in most basins in the central/western US. We found, except in the Northeast and mid-Atlantic basins, a decrease in the seasonality index, indicating floods arriving more uniformly is typically associated with an increase in the precipitation days in basins. On the other hand, increase in the seasonality index, indicating floods occurring more concentrated in time, is typically associated with an increase in the extreme precipitation in basins. Among the basin characteristics, elevation has a more dominant role than the drainage area in changing the flood seasonality. Elevation affects the form of precipitation, particularly in the western US, because floods arrive more distributed over the year (i.e., decrease in flood seasonality index), which potentially indicates increased warming resulting in early snowmelt.

Toward structuring real-world data: Deep learning for extracting oncology information from clinical text with patient-level supervision

Most detailed patient information in real-world data (RWD) is only consistently available in free-text clinical documents. Manual curation is expensive and time consuming. Developing natural language processing (NLP) methods for structuring RWD is thus essential for scaling real-world evidence generation. We propose leveraging patient-level supervision from medical registries, which are often readily available and capture key patient information, for general RWD applications. We conduct an extensive study on 135,107 patients from the cancer registry of a large integrated delivery network (IDN) comprising healthcare systems in five western US states. Our deep-learning methods attain test area under the receiver operating characteristic curve (AUROC) values of 94%-99% for key tumor attributes and comparable performance on held-out data from separate health systems and states. Ablation results demonstrate the superiority of these advanced deep-learning methods. Error analysis shows that our NLP system sometimes even corrects errors in registrar labels.