Next Article in Journal
Spatiotemporal Analysis and Automated Identification of Manning’s Roughness Coefficient in Lined Canals
Previous Article in Journal
A Physics-Based Method for Delineating Homogeneous Channel Units in Debris Flow Channels
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 17
Issue 23
10.3390/w17233445
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Ephemeral Channel Expansion: Predicting Shifts Toward Intermittency in Vulnerable Streams Across Semi-Arid CONUS

by
Lea J. Davidson
and
Adam M. Milewski
*
Water Resources & Remote Sensing Laboratory (WRRS), Department of Geology, University of Georgia, Athens, GA 30602, USA
*
Author to whom correspondence should be addressed.
Water 2025, 17(23), 3445; https://doi.org/10.3390/w17233445
Submission received: 10 November 2025 / Revised: 2 December 2025 / Accepted: 3 December 2025 / Published: 4 December 2025
(This article belongs to the Section Water and Climate Change)
Browse Figures
Versions Notes

Abstract

Broad trends point to the slow drying of streams, with warming temperatures and altered precipitation fueling declines in discharge across the Western United States. Sustained reductions in streamflow have the potential to drive the expansion of non-perennial channel networks, yet this process remains poorly characterized, with limited understanding of the variables which control stream vulnerability to intermittency or the spatial and temporal extent of these shifts. This research identifies significant trends toward novel intermittency across semi-arid regions of CONUS from 1980 to 2024. Of the 483 stream gages analyzed, more than half demonstrated reductions in discharge and increases in the frequency and duration of flow cessation. The relationship between flow intermittency and physical, hydrologic, climatic, and agricultural variables was further explored through discriminant function analysis (DFA). The timing of wet-season moisture, specifically December and January precipitation, was identified as the primary factor controlling the development of intermittency in semi-arid zones. With forecasted reductions in precipitation across CONUS, many currently perennial systems are vulnerable to developing intermittency. As a result, intermittent flow regimes are projected to expand further into previously perennial streams, as well as exacerbate dry-down across vulnerable channels.

1. Introduction

River systems are principal indicators of climate change, as they both reflect and respond to altered precipitation patterns and land surface temperatures. Stream networks across the landscape capture local and regional shifts in climate patterns and communicate these changes through variations in streamflow. Across the contiguous United States (CONUS), broad trends point to declines in stream discharge, particularly in the Southwest [1,2,3,4,5,6,7]. Drying rivers are, in part, a response to a steady reduction in total precipitation across dryland zones [8,9]. Trends in elevated land surface temperature and water cycle intensity further augment these declines by increasing evapotranspiration and reducing stored soil moisture and surface water [8,9,10,11]. These climate patterns mirror broader global trends, in which dryland regions are projected to shift toward increased aridity, altering local and regional river systems in the process [8,11,12,13].
Semi-arid regions encompass one-quarter of the CONUS land area, with 42% of drylands concentrated in the Western U.S. [13]. Streamflow in dryland river systems is distinctly seasonal. In line with precipitation, lowest flows are observed during the summer dry season, with the highest discharge during the wetter winter [1,8,9,14]. In the absence of consistent precipitation, perennial baseflow is sustained by groundwater discharge and headwater snowmelt [15,16,17]. For semi-arid streams with already limited discharge, reduced precipitation inputs have the potential to drive stream drying to the point of periodic disconnection and the development of intermittent flow regimes [18,19,20,21,22]. Though non-perennial and ephemeral streams are a significant and naturally occurring component of the global river network, their number is only expected to grow with increased aridity [19,23,24]. Limited attention has focused on present trends in the evolution of intermittent flow regimes or their projected future distribution, particularly in dryland perennial streams with the greatest vulnerability to reduced stream permanence [15,25]. Understanding and quantifying the factors which predispose a stream to shift from perennial to intermittent flow is crucial for sustainable water management, but complicated by interactions between numerous physical, climatic, and anthropogenic controls [5,7,19,20,26,27]. Aridity has been strongly linked to increased flow disruption; however, the dominant influence of climate has the potential to obscure the impact of other relevant variables in global or regional multi-climate zone studies [5,27,28,29].
Currently, the majority of river systems in the Western U.S. are managed under the assumption of the perennial availability of flow, with water users reliant on surface water to sustain agriculture through seasonally variable precipitation [6,30,31]. Failure to appropriately predict and account for developing stream intermittency has the potential to exacerbate stream drying and disconnection [31,32]. The anticipated alteration of historic flow patterns requires a shift from passive to proactive management of developing ephemerality, to ensure a future of sustainable water availability [33].
This research explores the development of novel stream intermittency and the primary variables driving shifts in historic flow from perennial to intermittent patterns. We conducted a large-scale analysis of developing intermittent flow regimes across 483 stream gages located in semi-arid zones of CONUS, for a minimum 30-year period from 1980 to 2024. Each gage was evaluated for trends in seasonal discharge, with gages identified as non-perennial assessed for the seasonal distribution and duration of zero-flow days. Physical, hydrologic, climatic, and anthropogenic variables were further compiled for each gage, with relationships evaluated via principal component analysis (PCA) and combined with discriminant function analysis (DFA) for flow differentiation and prediction. The goals of our analyses were to (1) characterize the current patterns and distribution of seasonal non-perennial flow across streams in semi-arid CONUS, (2) identify the dominant variables driving perennial vs. intermittent stream flow in semi-arid systems, and (3) predict a streams vulnerability to the development of flow intermittency in basins with limited gaging.

2. Materials and Methods

Daily measurements of stream discharge were sourced from 483 U.S. Geologic Survey (USGS) stream gages located within semi-arid zones of CONUS from 1980 to 2024. This time period was selected to maximize the number of gages with a continuous record. Gages were individually analyzed for seasonal trends in discharge, days with zero-stream flow, and the length of zero-flow periods. Each gage was further associated with 33 distinct variables related to stream intermittency for exploration of potential factors controlling the development of non-perennial flow regimes via linear discriminant function analysis (DFA). The resulting linear function was further used to predict flow intermittency in 448 stream gages with an insufficient data history for analysis, as well as identify the primary variables which predispose a stream section to non-perennial flow.

2.1. Gage Selection

Stream gages located within semi-arid regions of CONUS were the primary focus of this study, as dryland systems with both seasonal and limited precipitation have increased vulnerability to periodic flow cessation [15,34]. Semi-arid zones were defined by the Köppen–Geiger climate classification, which differentiates semi-arid steppe into subgroups based on mean annual air temperature (BSh and BSk) [13].
Active stream gage data was sourced from the USGS National Water Information System (NWIS) database [35]. USGS stream gages located within the defined semi-arid regions were filtered and further evaluated for sufficient daily discharge using R statistical software (v4.1.1) and the dataRetrieval R package (v2.7.12) [36,37]. Gages that lacked daily discharge values for a minimum of a continuous 30-year period between 1980 and 2024 were excluded, resulting in 483 gages with a sufficient data history for statistical analysis. Gages excluded for insufficient discharge data were filtered by unique sub-basins (USGS-defined 8-digit hydrologic unit), to produce a secondary dataset of 448 stream gages with unknown intermittency patterns [38].

2.2. Trend Analysis: Stream Drying

Seasonal Mann–Kendall trend analysis was applied to stream gages with sufficient data to identify seasonal trends in discharge and historic patterns of intermittent streamflow. In addition to seasonal trends in stream discharge (increasing, decreasing, or no change), two hydrologic signatures linked to intermittency were explored: zero-flow occurrence and zero-flow duration [27]. Seasonal zero-flow occurrence specifically estimates the trend in ‘no flow’ days occurring during the wet and dry seasons, with an increasing trend indicating a drier system with more zero-flow days. Seasonal zero-flow duration estimates the trend in the length of a no-flow period (the number of consecutive days without streamflow) over the same seasonal breakdown, with an increasing trend indicative of longer stretches in which the gage reads zero. For semi-arid CONUS, we broadly define the wet season as October through March, and the dry season as April through September [8,9]. Seasonal zero-flow occurrence was further used to classify each gage as either intermittent or perennial based on the presence of more than one period of consecutive zero-flow days.
Mann–Kendall trend analysis is a non-parametric statistical test used to estimate statistically significant trends in time series data [39,40,41]. In seasonal Mann–Kendall analysis, the trend for each season is calculated individually. This method is ideal for hydrologic data, which generally displays seasonal patterns related to precipitation and groundwater abstraction [41,42]. Seasonal discharge data was delineated monthly, while zero-flow occurrence and duration were calculated based on aggregated values within the broadly defined dry and wet seasons for North America [9]. Statistical analysis was performed using the R package Kendall (v.2.2.1) [43]. It is important to note that a gage reading of zero, though intended to signal flow cessation, may in fact represent equipment error, flow reversal, or management diversion [44]. To address this, stream gage data was further filtered for accuracy and quality assurance by the reporting agency.

2.3. Variable Selection and Data Collection

Diverse physical, climatic, and anthropogenic variables have been identified in previous studies as potentially significant factors controlling flow intermittency [5,20,27,34,45]. Thirty-three variables were selected for analysis and synthesized for each gage. All gages within semi-arid regions of CONUS, including both those with sufficient and insufficient data for analysis of intermittency, were selected, resulting in 30,723 measurements across 931 gages (Table 1).

2.4. Discriminant Function Analysis and Principal Component Analysis

Linear discriminant function analysis (DFA) was applied to identify variables with the strongest relationship to channel intermittency and to generate a linear function for prediction of non-perennial flow in systems with a limited gaging history. Linear DFA is a statistical method used to sort unknown continuous data into known categorical groups (here, intermittent or perennial channels) through the generation of a linear function built on provided variables [57]. This linear function maximizes the distinction between the known groups, utilizing the unique input variables for each gage to determine group membership as a classification probability. Perennial and non-perennial systems represent distinct categories; however, the lack of in situ gage data limits our ability to identify a system’s flow regime. The linear function generated through DFA enables accurate categorization, generating a probability of group membership for each gage based on a site’s unique variables. DFA is ideal for observed data with multiple predictor variables, allowing for rapid and flexible analysis without overfitting. Prior to analysis, a subset of variables was log transformed to normalize their data distribution. Z-scores were further calculated for all variables to standardize data for comparison. Stream gages with sufficient time series data were used as a training dataset, with the predictive accuracy of the linear DFA evaluated by whether its classifications matched known group membership [58]. Validation was performed through jackknife resampling which produces a secondary predictive accuracy, free of resubstitution error [58]. Following validation, the DFA linear function was applied to the secondary gage dataset with insufficient time series data to predict group membership and identify variables correlated to increased vulnerability to intermittent flow. DFA statistical analysis and data processing was performed in R using the MASS R package (v7.3–60) [59].
Principal component analysis (PCA) was additionally applied to explore variable relationships and corroborate observations from DFA. PCA is a statistical technique used to reduce the dimensionality of a dataset, particularly ideal for situations with many variables [60]. This method organizes data by reorienting axes based on decreasing variance [60]. PCA was performed on stream gages with sufficient time series data; a correlation biplot was additionally generated to visualize the spatial relationships between variables.

3. Results

3.1. Trends in Stream Drying

Stream gages distributed across semi-arid regions of CONUS reveal significant patterns of drying. From 1980 to 2024, 63.1% of stream gages demonstrated statistically significant seasonal trends in decreased discharge. Across the 483 gages analyzed, less than one-third (29.4%) experienced increases in seasonal discharge, while 7.5% of gages exhibited no statistically significant change. Though drying streams are distributed across all semi-arid regions of CONUS, they are most concentrated in the Southwest. In contrast, gages with increased discharge were focused in the Northern Great Plains (Figure 1a). Seasonal discharge trends were further explored for two signatures of stream intermittency: zero-flow occurrence and zero-flow duration. Seasonal zero-flow occurrence analyzes the quantity of zero-discharge days per season, while seasonal zero-flow duration explores the consecutive length of daily zero-flow gage readings. Periods of zero-flow occurrence were used to classify stream gages as either intermittent or perennial stream sections, with 48.4% of gages identified as non-perennial from 1980 to 2024. Statistically significant trends in zero-flow occurrence were observed in 118 stream gages identified as intermittent systems, with 60.2% experiencing an increase in the number of no-flow days per season (Figure 1b). Within these gages, 59.8% demonstrated a shift toward longer sustained periods without stream discharge, observed as increased trends in zero-flow duration. Stream gages undergoing a seasonal expansion in the frequency and length of zero-flow periods paralleled the spatial distribution of streams experiencing decreased seasonal discharge, with the greatest concentration in southwestern CONUS. Regardless of current flow regime, stream gages are broadly becoming drier, with 65.5% of perennial and 60.7% of intermittent gages demonstrating statistically significant seasonal reductions in flow. This implies the likely expansion of stream intermittency across CONUS, with novel intermittency developing in previous perennial systems and exacerbated intermittency in currently non-perennial channels.

3.2. Differentiating Drivers of Perennial and Intermittent Flow Regimes

In conjunction with projected shifts toward stream intermittency, primary drivers of non-perennial flow were identified for exclusively dryland zones. Dominant variables contributing to the distinction between perennial and intermittent systems were identified through linear discriminant function analysis (DFA). Trained on gages with known flow regime, the resulting linear DFA demonstrates high predictive accuracy, correctly classifying stream sections as perennial or intermittent with 79.5% accuracy (Table 2). Perennial and intermittent flow regimes are identified as distinct states by the linear DFA, which plotted supports discrete group classification with minimal overlap (Figure 2). From this linear function, variables related to wet-season moisture were identified as the greatest influence on differentiating perennial from intermittent stream flow (Figure 3). During October through March, the timing of precipitation and soil moisture variables at distinct points within the wet season primarily drove group separation. Average December precipitation had the overall greatest influence on distinguishing between stream types (Figure 3 and Figure 4). Specifically, average December precipitation and average soil moisture for both January and March were identified as the greatest controlling variables for intermittent streams (Figure 3 and Figure 4). Average precipitation for January and November, paired with average February soil moisture, demonstrated the greatest impact on perennial flow (Figure 3 and Figure 4). The timing of dominant precipitation variables implies that the concentration and maintenance of moisture at the end of the wet season is largely what separates an intermittent from a perennial system. Excluding climate variables, which demonstrate the greatest overall contribution to the DFA’s group differentiation, several physical variables exerted a secondary influence on intermittency. Specifically, the contributing drainage area of each gage, the dominance of both shrub/scrub land cover, and slow infiltration soils (class C) within each sub-basin were identified as non-climate related controls with influence on group differentiation (Figure 5). It is interesting to note that variables related to anthropogenic influence, specifically irrigated land area and number of dams within the sub-basin, demonstrated markedly low influence on differentiation of flow regime.
The role of moisture seasonality was further validated through PCA, which established an inverse relationship between wet-season precipitation and dry-season soil moisture (Figure 6). Performed on stream gages with sufficient time series data, the PCA generated four principal components which explained 78.1% of the data variance (Figure 6a). This relationship underscores the potential extended impact of concentrated wet-season moisture inputs on sustained antecedent moisture in dry-season months. Correlation between decreased wet-season soil moisture and increased elevation was further observed, suggesting the role of snowpack in higher elevation gages to sustain perennial flow during the lower precipitation dry season. This may suggest the potential for snow storage to buffer a system from novel ephemerality.
Across stream gages located in semi-arid regions of CONUS that lacked sufficient data history for trend analysis, the previous DFA characterization was applied to predict gage vulnerability to non-perennial flow. Of the 448 gages with insufficient discharge data, 49.6% were predicted to experience intermittent flow patterns (Figure 7). Gages predicted to be intermittent were uniformly spatially distributed across semi-arid CONUS, with minor concentration in the Southwest and Southern Great Plains. The exception was the upper Rocky Mountain region (Utah, Wyoming, Colorado), in which the majority of gages were predicted to be in perennial flowing systems.

4. Discussion

Our results highlight broad shifts in intermittent flow regimes across dryland CONUS, with increasing trends toward drying and expansion of intermittency. Stream gages document both an elevated frequency of zero-flow periods and progressively longer stretches before water returns. Similar trends in streamflow have been identified in previous work across CONUS; however, these findings fail to characterize the nuance of climate variables driving these shifts, outside of general aridity [27]. This work identifies the timing and distribution of moisture inputs, particularly within the winter wet season, as the primary control on non-perennial flow within exclusively semi-arid zones. Soil moisture is closely tied to precipitation in the month preceding, and higher precipitation in the latter half of the wet season (December and January) is expected to translate to elevated soil moisture and overall increased water storage within a hydrologic system [61,62]. Accumulation of moisture, particularly as the wet season draws to a close, may likely help to sustain discharge throughout the advancing dry season. It is important to note that the relationship between December precipitation/intermittent flow regimes and January precipitation/perennial flow regimes makes no assumption of elevated rainfall as important for group distinction. We instead theorize that comparably low levels of December precipitation are likely a significant differentiator in intermittent channels, while higher levels of January precipitation are associated with perennial flow. Our results further highlight the importance of early-onset wet season precipitation for controlling channel intermittency. The initial rains in November generally mark a transition to the wet season, bringing with them not only moisture inputs, which may reactivate a non-flowing or low-discharge channel, but ushering in a drop in temperature and reduced evapotranspiration. Soil moisture content plays a significant role in arid systems, where antecedent sediment moisture is necessary for sustained infiltration to the water table [63]. Early wet-season precipitation can drive increased antecedent soil moisture and percolation resulting in overall greater seasonal recharge, through both focused recharge via the stream channel and diffuse recharge across the landscape [61]. Development of sustained antecedent soil moisture early in the season may allow for smaller precipitation events to be translated to recharge instead of evaporation. This has the potential to raise the height of the water table and increase groundwater discharge into streams, supporting perennial baseflow.
Viewed together, this analysis underscores the impact of bracketed seasonal moisture on the development of channel intermittency, defined as both precipitation and soil moisture at the onset and conclusion of the wet season. Bracketed seasonal moisture may be a primary control on total wet season recharge, as it supports ideal conditions for greater total recharge and primes a system for elevated baseflow during the transition to the dry season. Outcomes from this work suggest that precipitation timing, particularly within the wet season, may be a stronger predictor of developing intermittency than general climate variables such as total annual precipitation or evapotranspiration.
Projected climate patterns across semi-arid regions of CONUS are, however, forecast to increase the seasonality of precipitation, with a reduction in the frequency of storms and an increase in the size of singular events [64,65]. This shift may impact the timing of precipitation, which has the potential to both delay wet-season onset and expedite its conclusion [65]. Altered timing of precipitation may impact the continuity of antecedent soil moisture and ultimately the volume of water able to effectively infiltrate and replenish groundwater stores during the wet season. This has the potential to contribute to the increased development of intermittent flow regimes. In addition, present patterns in water cycle intensity are projected to shift across CONUS, with declines in total precipitation and stable evapotranspiration (Figure 8) [10]. This scenario could likely exacerbate projected stream drying, increasing both the distribution and acute severity within individual sections. Decreases in water cycle intensity lay the groundwork for an even greater expansion of intermittency into historically perennial stream reaches, including higher elevation sub-regions of the Mountain West.
Regional water management strategies, such as artificially maintained stream flow, diversions, or adjacent pumping, have the potential either to slow or exacerbate predicted drying trends. Though this analysis did not identify a significant impact of anthropogenic influence on the differentiation of flow regime, observed trends in stream discharge may be intensified by regional water management practices. This influence should be further explored at the sub-regional level, as management decision-making and needs can be highly localized. Future work which explores the impact of various management strategies on developing intermittency is important to improve predictions of expanding ephemerality under future climate scenarios.
Expanding ephemerality has important implications not only for hydrologic, but physical and ecologic processes, with the potential to disturb aquatic communities, sediment transport, and nutrient flux [23,66]. It further has significant consequences for sustainable water management. Though current management may consider seasonal variability in stream discharge, this thinking must be expanded to include periods of stream drought. Failure to manage zero-discharge intervals has the potential to exacerbate stream dry downs, increasing their duration and stressing alternative water sources such as regional groundwater. Insufficient preparation for novel non-perennial flow will likely have a substantial economic impact, particularly on regional agriculture which relies on stream flow to support irrigation during the dry summer months [31].

5. Conclusions

Broad shifts towards decreased stream discharge across semi-arid CONUS highlight the necessity of efficient characterization for stream management and conservation. Previously perennial streams are increasingly shifting toward novel non-perennial flow, while intermittent stream channels demonstrate increased duration and occurrence of no-flow periods. The timing of moisture inputs, particularly at the onset and conclusion of the wet season, were identified as dominant controls of developing stream intermittency. Paired with projected shifts in regional climate patterns and water cycle intensity, the distribution of non-perennial systems across semi-arid CONUS is only expected to expand. Improved understanding of the factors which predispose a stream to novel intermittency are crucial for improved management and prediction of expanding ephemerality, both domestically and in vulnerable dryland basins around the globe.

Author Contributions

L.J.D. was responsible for research question conceptualization, investigation, analysis, writing, and editing. A.M.M. supported project administration, supervision, conceptualization, review, and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the University of Georgia, Department of Geology funding L.J.D. and A.M.M.

Data Availability Statement

Data presented in this study are available on request from the corresponding author. All stream gage data is sourced from the USGS Water Data for the Nation (https://waterdata.usgs.gov, accessed on 1 April 2024). All stream gage data, satellite estimates, and processing software are publicly available.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Lins, H.F.; Slack, J.R. Seasonal and Regional Characteristics of U.S. Streamflow Trends in the United States from 1940 to 1999. Phys. Geogr. 2005, 26, 489–501. [Google Scholar] [CrossRef]
  2. Ficklin, D.L.; Robeson, S.M.; Knouft, J.H. Impacts of Recent Climate Change on Trends in Baseflow and Stormflow in United States Watersheds. Geophys. Res. Lett. 2016, 43, 5079–5088. [Google Scholar] [CrossRef]
  3. Dudley, R.W.; Hirsch, R.M.; Archfield, S.A.; Blum, A.G.; Renard, B. Low Streamflow Trends at Human-Impacted and Reference Basins in the United States. J. Hydrol. 2020, 580, 124254. [Google Scholar] [CrossRef]
  4. Dethier, E.N.; Sartain, S.L.; Renshaw, C.E.; Magilligan, F.J. Spatially Coherent Regional Changes in Seasonal Extreme Streamflow Events in the United States and Canada since 1950. Sci. Adv. 2020, 6, eaba5939. [Google Scholar] [CrossRef] [PubMed]
  5. Hammond, J.C.; Zimmer, M.; Shanafield, M.; Kaiser, K.; Godsey, S.E.; Mims, M.C.; Zipper, S.C.; Burrows, R.M.; Kampf, S.K.; Dodds, W.; et al. Spatial Patterns and Drivers of Nonperennial Flow Regimes in the Contiguous United States. Geophys. Res. Lett. 2021, 48, e2020GL090794. [Google Scholar] [CrossRef]
  6. Zipper, S.C.; Farmer, W.H.; Brookfield, A.; Ajami, H.; Reeves, H.W.; Wardropper, C.; Hammond, J.C.; Gleeson, T.; Deines, J.M. Quantifying Streamflow Depletion from Groundwater Pumping: A Practical Review of Past and Emerging Approaches for Water Management. J. Am. Water Resour. Assoc. 2022, 58, 289–312. [Google Scholar] [CrossRef]
  7. Hammond, J.C.; Simeone, C.; Hecht, J.S.; Hodgkins, G.A.; Lombard, M.; McCabe, G.; Wolock, D.; Wieczorek, M.; Olson, C.; Caldwell, T.; et al. Going Beyond Low Flows: Streamflow Drought Deficit and Duration Illuminate Distinct Spatiotemporal Drought Patterns and Trends in the U.S. During the Last Century. Water Resour. Res. 2022, 58, e2022WR031930. [Google Scholar] [CrossRef]
  8. Seager, R.; Neelin, D.; Simpson, I.; Liu, H.; Henderson, N.; Shaw, T.; Kushnir, Y.; Ting, M.; Cook, B. Dynamical and Thermodynamical Causes of Large-Scale Changes in the Hydrological Cycle over North America in Response to Global Warming. J. Clim. 2014, 27, 7921–7948. [Google Scholar] [CrossRef]
  9. Ting, M.; Seager, R.; Li, C.; Liu, H.; Henderson, N. Mechanism of Future Spring Drying in the Southwestern United States in CMIP5 Models. J. Clim. 2018, 31, 4265–4279. [Google Scholar] [CrossRef]
  10. Zowam, F.J.; Milewski, A.M.; Richards IV, D.F. A Satellite-Based Approach for Quantifying Terrestrial Water Cycle Intensity. Remote Sens. 2023, 15, 3632. [Google Scholar] [CrossRef]
  11. U.S. Environmental Protection Agency. Climate Change Indicators in the United States, 5th ed.; EPA 430-R-24-003; U.S. Environmental Protection Agency: Washington, DC, USA, 2024.
  12. Scheff, J.; Frierson, D.M.W. Robust Future Precipitation Declines in CMIP5 Largely Reflect the Poleward Expansion of Model Subtropical Dry Zones. Geophys. Res. Lett. 2012, 39, 18. [Google Scholar] [CrossRef]
  13. Beck, H.E.; Zimmermann, N.E.; McVicar, T.R.; Vergopolan, N.; Berg, A.; Wood, E.F. Present and Future Köppen-Geiger Climate Classification Maps at 1-Km Resolution. Sci. Data 2018, 5, 180214. [Google Scholar] [CrossRef] [PubMed]
  14. Goodrich, D.C.; Kepner, W.G.; Levick, L.R.; Wigington, P.J., Jr. Southwestern Intermittent and Ephemeral Stream Connectivity. JAWRA J. Am. Water Resour. Assoc. 2018, 54, 400–422. [Google Scholar] [CrossRef]
  15. Levick, L.; Fonseca, J.; Goodrich, D.; Hernandez, M.; Semmens, D.; Stromberg, J.; Leidy, R.; Scianni, M.; Guertin, D.P.; Tluczek, M.; et al. The Ecological and Hydrological Significance of Ephemeral and Intermittent Streams in the Arid and Semi-Arid American Southwest; U.S. Environmental Protection Agency Office of Research and Development: Washington, DC, USA, 2008.
  16. Gleeson, T.; Cuthbert, M.; Ferguson, G.; Perrone, D. Global Groundwater Sustainability, Resources, and Systems in the Anthropocene. Annu. Rev. Earth Planet. Sci. 2020, 48, 431–463. [Google Scholar] [CrossRef]
  17. Chen, X.; Tang, G.; Chen, T.; Niu, X. An Assessment of the Impacts of Snowmelt Rate and Continuity Shifts on Streamflow Dynamics in Three Alpine Watersheds in the Western U.S. Water 2022, 14, 1095. [Google Scholar] [CrossRef]
  18. Jaeger, K.L.; Olden, J.D.; Pelland, N.A. Climate Change Poised to Threaten Hydrologic Connectivity and Endemic Fishes in Dryland Streams. Proc. Natl. Acad. Sci. USA 2014, 111, 13894–13899. [Google Scholar] [CrossRef]
  19. Costigan, K.H.; Jaeger, K.L.; Goss, C.W.; Fritz, K.M.; Goebel, P.C. Understanding Controls on Flow Permanence in Intermittent Rivers to Aid Ecological Research: Integrating Meteorology, Geology and Land Cover: Integrating Science to Understand Flow Intermittence. Ecohydrology 2016, 9, 1141–1153. [Google Scholar] [CrossRef]
  20. Schilling, O.S.; Cook, P.G.; Grierson, P.F.; Dogramaci, S.; Simmons, C.T. Controls on Interactions Between Surface Water, Groundwater, and Riverine Vegetation Along Intermittent Rivers and Ephemeral Streams in Arid Regions. Water Resour. Res. 2021, 57, e2020WR028429. [Google Scholar] [CrossRef]
  21. Sauquet, E.; Shanafield, M.; Hammond, J.C.; Sefton, C.; Leigh, C.; Datry, T. Classification and Trends in Intermittent River Flow Regimes in Australia, Northwestern Europe and USA: A Global Perspective. J. Hydrol. 2021, 597, 126170. [Google Scholar] [CrossRef]
  22. Kelly, B.T.; Bruckerhoff, L.A. Dry, Drier, Driest: Differentiating Flow Patterns across a Gradient of Intermittency. River Res. Appl. 2024, 40, 1273–1285. [Google Scholar] [CrossRef]
  23. Datry, T.; Larned, S.T.; Tockner, K. Intermittent Rivers: A Challenge for Freshwater Ecology. BioScience 2014, 64, 229–235. [Google Scholar] [CrossRef]
  24. Messager, M.L.; Lehner, B.; Cockburn, C.; Lamouroux, N.; Pella, H.; Snelder, T.; Tockner, K.; Trautmann, T.; Watt, C.; Datry, T. Global Prevalence of Non-Perennial Rivers and Streams. Nature 2021, 594, 391–397. [Google Scholar] [CrossRef]
  25. Zimmer, M.A.; Burgin, A.J.; Kaiser, K.; Hosen, J. The Unknown Biogeochemical Impacts of Drying Rivers and Streams. Nat. Commun. 2022, 13, 7213. [Google Scholar] [CrossRef]
  26. Jaeger, K.; Sando, R.; McShane, R.R.; Dunham, J.B.; Hockman-Wert, D.; Kaiser, K.; Hafen, K.; Risley, J.C.; Blasch, K. Probability of Streamflow Permanence Model (PROSPER): A Spatially Continuous Model of Annual Streamflow Permanence throughout the Pacific Northwest. J. Hydrol. X 2018, 2, 100005. [Google Scholar] [CrossRef]
  27. Zipper, S.C.; Hammond, J.C.; Shanafield, M.; Zimmer, M.; Datry, T.; Jones, C.N.; Kaiser, K.E.; Godsey, S.E.; Burrows, R.M.; Blaszczak, J.R.; et al. Pervasive Changes in Stream Intermittency across the United States. Environ. Res. Lett. 2021, 16, 084033. [Google Scholar] [CrossRef]
  28. Eng, K.; Wolock, D.M.; Dettinger, M.D. Sensitivity of Intermittent Streams to Climate Variations in the USA. River Res. Appl. 2016, 32, 885–895. [Google Scholar] [CrossRef]
  29. Peña-Gallardo, M.; Vicente-Serrano, S.M.; Hannaford, J.; Lorenzo-Lacruz, J.; Svoboda, M.; Domínguez-Castro, F.; Maneta, M.; Tomas-Burguera, M.; Kenawy, A.E. Complex Influences of Meteorological Drought Time-Scales on Hydrological Droughts in Natural Basins of the Contiguous Unites States. J. Hydrol. 2019, 568, 611–625. [Google Scholar] [CrossRef]
  30. Kampf, S.K.; Dwire, K.A.; Fairchild, M.P.; Dunham, J.; Snyder, C.D.; Jaeger, K.L.; Luce, C.H.; Hammond, J.C.; Wilson, C.; Zimmer, M.A.; et al. Managing Nonperennial Headwater Streams in Temperate Forests of the United States. For. Ecol. Manag. 2021, 497, 119523. [Google Scholar] [CrossRef]
  31. Ketchum, D.; Hoylman, Z.H.; Huntington, J.; Brinkerhoff, D.; Jensco, K.G. Irrigation Intensification Impacts Sustainability of Streamflow in the Western United States. Commun. Earth Environ. 2023, 4, 479. [Google Scholar] [CrossRef]
  32. O’Connor, B.L.; Hamada, Y.; Bowen, E.E.; Grippo, M.A.; Hartmann, H.M.; Patton, T.L.; Van Lonkhuyzen, R.A.; Carr, A.E. Quantifying the Sensitivity of Ephemeral Streams to Land Disturbance Activities in Arid Ecosystems at the Watershed Scale. Environ. Monit Assess 2014, 186, 7075–7095. [Google Scholar] [CrossRef]
  33. Kovach, R.; Dunham, J.; Al-Chokhachy, R.; Snyder, C.; Letcher, B.; Young, J.; Beever, E.; Pederson, G.; Lynch, A.; Hitt, N.; et al. An Integrated Framework for Ecological Drought across Riverscapes of North America. BioScience 2019, 69, 418–431. [Google Scholar] [CrossRef]
  34. Boulton, A.J.; Rolls, R.J.; Jaeger, K.L.; Datry, T. Chapter 2.3—Hydrological Connectivity in Intermittent Rivers and Ephemeral Streams. In Intermittent Rivers and Ephemeral Streams; Datry, T., Bonada, N., Boulton, A., Eds.; Academic Press: Cambridge, MA, USA, 2017; pp. 79–108. ISBN 978-0-12-803835-2. [Google Scholar]
  35. U.S. Geological Survey USGS. National Water Information System (NWIS); Water Data for the Nation: Reston, VA, USA, 2001.
  36. DeCicco, L. Water/dataRetrieval·GitLab. Available online: https://code.usgs.gov/water/dataRetrieval (accessed on 20 August 2024).
  37. R Core Team. R: A Language and Environment for Statistical Computing; R Core Team: Vienna, Austria, 2011. [Google Scholar]
  38. U.S. Geological Survey and U.S. Department of Agriculture; Natural Resources Conservation Service. Federal Standards and Procedures for the National Watershed Boundary Dataset (WBD), 4th ed.; U.S. Geological Survey, Techniques and Methods 11–A3; U.S. Geological Survey: Reston, VA, USA, 2013; p. 63.
  39. Mann, H.B. Nonparametric Tests Against Trend. Econometrica 1945, 13, 245–259. [Google Scholar] [CrossRef]
  40. Kendall, M.G. Rank Correlation Methods; C. Griffin: Spokane, WA, USA, 1948. [Google Scholar]
  41. Meals, D.M.; Spooner, J.; Dressing, S.A.; Harcum, J.B. Statistical Analysis for Monotonic Trends; U.S. Environmental Protection Agency by Tetra Tech, Inc.: Fairfax, VA, USA, 2011.
  42. Hirsch, R.M.; Slack, J.R.; Smith, R.A. Techniques of Trend Analysis for Monthly Water Quality Data. Water Resour. Res. 1982, 18, 107–121. [Google Scholar] [CrossRef]
  43. McLeod, A.I. Kendall: Kendall Rank Correlation and Mann-Kendall Trend Test. R Package Version 2.2.1. 2022. Available online: https://CRAN.R-Project.Org/package=Kendall (accessed on 1 April 2024).
  44. Zimmer, M.A.; Kaiser, K.E.; Blaszczak, J.R.; Zipper, S.C.; Hammond, J.C.; Fritz, K.M.; Costigan, K.H.; Hosen, J.; Godsey, S.E.; Allen, G.H.; et al. Zero or Not? Causes and Consequences of Zero-flow Stream Gage Readings. WIREs Water 2020, 7, e1436. [Google Scholar] [CrossRef]
  45. Costigan, K.H.; Daniels, M.D.; Dodds, W.K. Fundamental Spatial and Temporal Disconnections in the Hydrology of an Intermittent Prairie Headwater Network. J. Hydrol. 2015, 522, 305–316. [Google Scholar] [CrossRef]
  46. U.S. Department of Transportation (USDOT). National Inventory of Dams; U.S. Department of Transportation (USDOT): Washington, DC, USA, 2013.
  47. Texas Commission on Environmental Quality (TCEQ). State Regulated Dams; Texas Commission on Environmental Quality (TCEQ): Austin, TX, USA, 2021.
  48. Elnashar, A.; Wang, L.; Wu, B.; Zhu, W.; Zeng, H. Synthesis of Global Actual Evapotranspiration from 1982 to 2019. Earth Syst. Sci. Data 2021, 13, 447–480. [Google Scholar] [CrossRef]
  49. Martin, D.J.; Regan, R.S.; Haynes, J.V.; Read, A.L.; Henson, W.R.; Stewart, J.S.; Brandt, J.T.; Niswonger, R.G. Irrigation Water Use Reanalysis for the 2000-20 Period by HUC12, Month, and Year for the Conterminous United States; U.S. Geological Survey Data Release: Sacramento, CA, USA, 2023.
  50. Dewitz, J. U.S. Geological Survey National Land Cover Database (NLCD) 2019 Land Cover Science Product; version 2.0; U.S. Geological Survey Data Release: Sacramento, CA, USA, 2021. [CrossRef]
  51. Zhang, T.; Zhou, Y.; Zhu, Z.; Li, X.; Asrar, G.R. A Global Seamless 1 Km Resolution Daily Land Surface Temperature Dataset (2003–2020). Earth Syst. Sci. Data 2021, 14, 651–664. [Google Scholar] [CrossRef]
  52. PRISM Climate Group; Oregon State University. 30-Year Normal Precipitation: Monthly 2014. Available online: https://prism.oregonstate.edu (accessed on 1 April 2024).
  53. Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture. Soil Survey Geographic (SSURGO) Database for CONUS. Available online: https://www.nrcs.usda.gov/resources/data-and-reports/soil-survey-geographic-database-ssurgo (accessed on 1 April 2024).
  54. Hongtao, J.; Huanfeng, S.; Xinghua, L.; Lili, L. The 43-Year (1978–2020) Global 9km Remotely Sensed Soil Moisture Product 2022; PANGAEA: Bremen, Germany, 2022. [Google Scholar]
  55. Zowam, F.J.; Milewski, A.M. Groundwater Level Prediction Using Machine Learning and Geostatistical Interpolation Models. Water 2024, 16, 2771. [Google Scholar] [CrossRef]
  56. U.S. Geological Survey. 1 Arc-Second Digital Elevation Models (DEMs)—USGS National Map 3DEP Downloadable Data Collection; U.S. Geological Survey: Reston, VA, USA, 2022.
  57. Davis, J.C. Statistics and Data Analysis in Geology, 3rd ed.; Wiley: New York, NY, USA, 2002. [Google Scholar]
  58. Maindonald, J.; Braun, W.J. Data Analysis and Graphics Using R—An Example-Based Approach, 3rd ed.; Cambridge University Press: Cambridge, UK, 2003; ISBN 978-0-511-71286-9. [Google Scholar]
  59. Venables, W.N.; Ripley, B.D. Modern Applied Statistics with S, 4th ed.; Springer: New York, NY, UA, 2002; ISBN 0-387-95457-0. [Google Scholar]
  60. Lever, J.; Krzywinski, M.; Altman, N. Principal Component Analysis. Nat. Methods 2017, 14, 641–642. [Google Scholar] [CrossRef]
  61. Small, E.E. Climatic Controls on Diffuse Groundwater Recharge in Semiarid Environments of the Southwestern United States. Water Resour. Res. 2005, 41, W04012. [Google Scholar] [CrossRef]
  62. Sehler, R.; Li, J.; Reager, J.; Ye, H. Investigating Relationship Between Soil Moisture and Precipitation Globally Using Remote Sensing Observations. J. Contemp. Water Res. Educ. 2019, 168, 106–118. [Google Scholar] [CrossRef]
  63. Fakir, Y.; Bouimouass, H.; Constantz, J. Seasonality in Intermittent Streamflow Losses Beneath a Semiarid Mediterranean Wadi. Water Resour. Res. 2021, 57, e2021WR029743. [Google Scholar] [CrossRef]
  64. Chang, W.; Stein, M.L.; Wang, J.; Kotamarthi, V.R.; Moyer, E.J. Changes in Spatiotemporal Precipitation Patterns in Changing Climate Conditions. J. Clim. 2016, 29, 8355–8376. [Google Scholar] [CrossRef]
  65. Wasko, C.; Nathan, R.; Stein, L.; O’Shea, D. Evidence of Shorter More Extreme Rainfalls and Increased Flood Variability under Climate Change. J. Hydrol. 2021, 603, 126994. [Google Scholar] [CrossRef]
  66. Shumilova, O.; Zak, D.; Datry, T.; Schiller, D.; Corti, R.; Foulquier, A.; Obrador, B.; Tockner, K.; Allan, D.C.; Altermatt, F.; et al. Simulating Rewetting Events in Intermittent Rivers and Ephemeral Streams: A Global Analysis of Leached Nutrients and Organic Matter. Glob. Change Biol. 2019, 25, 1591–1611. [Google Scholar] [CrossRef] [PubMed]
Water 17 03445 g001
Figure 1. Drying streams across semi-arid CONUS. (a) Location and trend of 483 stream gages within semi-arid regions of CONUS (Köppen–Geiger zones highlighted in green and yellow). First-order streams outlined in gray. Statistically significant trends in seasonal discharge across stream gages were indicated by color, as declining stream discharge (orange), increasing stream discharge (blue), and no change in discharge trend (dark gray). A histogram depicts the overall distribution of gages with drying or wetting trends. Declining trends in discharge are particularly concentrated in the Southwest. (b) Across stream gages with intermittent flow patterns, periods of intermittency are increasing. Gage color symbolizes trend in zero-flow frequency, with gage size indicative of increasing or decreasing strength of trend. Overall, 60.2% of stream gages indicate an increase in the frequency of a completely dry channel. A histogram depicts the overall distribution of gage trends in zero-flow frequency.
Figure 1. Drying streams across semi-arid CONUS. (a) Location and trend of 483 stream gages within semi-arid regions of CONUS (Köppen–Geiger zones highlighted in green and yellow). First-order streams outlined in gray. Statistically significant trends in seasonal discharge across stream gages were indicated by color, as declining stream discharge (orange), increasing stream discharge (blue), and no change in discharge trend (dark gray). A histogram depicts the overall distribution of gages with drying or wetting trends. Declining trends in discharge are particularly concentrated in the Southwest. (b) Across stream gages with intermittent flow patterns, periods of intermittency are increasing. Gage color symbolizes trend in zero-flow frequency, with gage size indicative of increasing or decreasing strength of trend. Overall, 60.2% of stream gages indicate an increase in the frequency of a completely dry channel. A histogram depicts the overall distribution of gage trends in zero-flow frequency.
Water 17 03445 g001
Water 17 03445 g002
Figure 2. DFA. Graphical distribution of intermittent versus perennial streams based on DFA. The histogram depicts distinct grouping and tight within-group distribution indicative of effective group separation by the linear function. Some overlap between intermittent versus perennial stream groups may be expected, as gages which currently exist within a distinct group may be in the process of transitioning to the other.
Figure 2. DFA. Graphical distribution of intermittent versus perennial streams based on DFA. The histogram depicts distinct grouping and tight within-group distribution indicative of effective group separation by the linear function. Some overlap between intermittent versus perennial stream groups may be expected, as gages which currently exist within a distinct group may be in the process of transitioning to the other.
Water 17 03445 g002
Water 17 03445 g003
Figure 3. DFA loadings for 33 predictive variables. Loadings indicate the contribution of each variable to the predictor function. Larger values indicate greater overall contribution. This is further underscored by sorting and coloring within the figure, with variables sorted by decreasing level of contribution and colored by group influence (perennial streams in blue, intermittent streams in orange).
Figure 3. DFA loadings for 33 predictive variables. Loadings indicate the contribution of each variable to the predictor function. Larger values indicate greater overall contribution. This is further underscored by sorting and coloring within the figure, with variables sorted by decreasing level of contribution and colored by group influence (perennial streams in blue, intermittent streams in orange).
Water 17 03445 g003
Water 17 03445 g004
Figure 4. Timing of wet-season moisture factors controlling the development of intermittency. Average December precipitation exerts the most significant influence on non-perennial flow. Sustained moisture inputs late in the wet season are correlated to strong distinctions between perennial channels perennially and those which dry up.
Figure 4. Timing of wet-season moisture factors controlling the development of intermittency. Average December precipitation exerts the most significant influence on non-perennial flow. Sustained moisture inputs late in the wet season are correlated to strong distinctions between perennial channels perennially and those which dry up.
Water 17 03445 g004
Water 17 03445 g005
Figure 5. Non-climate variables exert minor influence on stream intermittency. Contributing basin drainage area (displayed as gage marker size), and sub-basins dominated by shrub/scrub (green) and slow infiltration soils (pink) display correlation to group distinction. Intuitively, perennial streams are observed to correlate to greater contributing drainage areas than intermittent systems. The contribution of shrub/scrub dominance to system intermittency may be a secondary identifier of the arid climate conditions associated with stream ephemerality. Dominant vegetation type proves to be a stronger contributor to group differentiation than seemingly more direct climate variables, such as land surface temperature or average evapotranspiration.
Figure 5. Non-climate variables exert minor influence on stream intermittency. Contributing basin drainage area (displayed as gage marker size), and sub-basins dominated by shrub/scrub (green) and slow infiltration soils (pink) display correlation to group distinction. Intuitively, perennial streams are observed to correlate to greater contributing drainage areas than intermittent systems. The contribution of shrub/scrub dominance to system intermittency may be a secondary identifier of the arid climate conditions associated with stream ephemerality. Dominant vegetation type proves to be a stronger contributor to group differentiation than seemingly more direct climate variables, such as land surface temperature or average evapotranspiration.
Water 17 03445 g005
Water 17 03445 g006
Figure 6. PCA. (a) Four principal components explain 78.1% of data variance, and (b) relationships between variables demonstrate distinct seasonal groupings.
Figure 6. PCA. (a) Four principal components explain 78.1% of data variance, and (b) relationships between variables demonstrate distinct seasonal groupings.
Water 17 03445 g006
Water 17 03445 g007
Figure 7. Many unclassified streams are vulnerable to intermittency. Across semi-arid CONUS, 49.6% of unknown gages are predicted to be highly vulnerable to stream intermittency. Minor concentrations are observed in the Southwest and Southern Great Plains. Potential for intermittent and perennial flow across CONUS is represented via the Esri Empirical Bayesian Kriging method.
Figure 7. Many unclassified streams are vulnerable to intermittency. Across semi-arid CONUS, 49.6% of unknown gages are predicted to be highly vulnerable to stream intermittency. Minor concentrations are observed in the Southwest and Southern Great Plains. Potential for intermittent and perennial flow across CONUS is represented via the Esri Empirical Bayesian Kriging method.
Water 17 03445 g007
Water 17 03445 g008
Figure 8. Stream intermittency is projected to expand. Changing climate patterns, particularly reductions in precipitation paired with stable ET (lower water cycle intensity), highlight the potential expansion of non-perennial flow into historically perennial stream reaches. This figure depicts gages with known intermittency patterns used to train the linear DFA and prediction of flow regime for gages with insufficient data history. Average water cycle intensity is adapted from Zowam et al. (2023) and based on gridded precipitation and ET datasets from 2001 to 2019 [10].
Figure 8. Stream intermittency is projected to expand. Changing climate patterns, particularly reductions in precipitation paired with stable ET (lower water cycle intensity), highlight the potential expansion of non-perennial flow into historically perennial stream reaches. This figure depicts gages with known intermittency patterns used to train the linear DFA and prediction of flow regime for gages with insufficient data history. Average water cycle intensity is adapted from Zowam et al. (2023) and based on gridded precipitation and ET datasets from 2001 to 2019 [10].
Water 17 03445 g008
Table 1. Variable Data Source and Availability. Variables with an asterisk were log transformed prior to analysis to normalize their distribution required for the statistical analysis.
Table 1. Variable Data Source and Availability. Variables with an asterisk were log transformed prior to analysis to normalize their distribution required for the statistical analysis.
VariablesData DescriptionSpatial Resolution
Contributing Drainage AreaUSGS NWIS dataset contributing drainage area is reported in sq. miles. A minority of stream gages lacked this metric, and USGS-designated watershed areas (HUC 10) were used as a substitute [35].30 m
Dams* USDOT National Inventory of Dams dataset. The state of Texas was supplemented with an inventory of state-regulated dams (TCEQ). Dam counts were aggregated for each gage sub-basin (HUC 8) [46,47].NA
ElevationUSGS-derived gage elevation (ft), from USGS NWIS dataset. Values were supplemented by elevation from USGS 1-arc second DEM [35].30 m
Evapotranspiration (ET)Multi-product and satellite-aggregated global ET dataset. ET was averaged monthly from 2010 to 2019, with average ET calculated for each gage sub-basin (HUC 8) [48].0.1°
Irrigated Area*Landsat-based irrigation dataset (LANID) for CONUS (2018–2020). Percent irrigated area was aggregated for each gage sub-basin (HUC 8) [49].30 m
Land Cover NLCD MRLC land cover classification for CONUS (2021). Dominant land cover type (percentage of total area) within gage sub-basin (HUC 8) was extracted [50].30 m
Maximum Land Surface Temperature (LST)NASA MODIS (Terra/Aqua)-generated dataset. LST averaged monthly from 2010 to 2019, with units converted to °F. Maximum LST within each gage sub-basin (HUC 8) was extracted [51].0.1°
Precipitation*PRISM-average monthly 30-year normal precipitation dataset (CONUS). Values were averaged across each gage sub-basin (HUC 8) [52].800 m
Soil Hydrologic GroupU.S. Soil Hydrologic Group (SSURGO) water infiltration classification. Dominant soil hydrologic group (percentage of total area) within gage sub-basin (HUC 8) was extracted [53].30 m
Soil Moisture (SM)*Multi-product and satellite-generated dataset, downscaled by Zowam and Milewski (2024) [54,55]. SM was averaged monthly from 2010 to 2019, and further averaged across each gage sub-basin (HUC 8).0.1°
SlopeAverage slope across each gage sub-basin (HUC 8), derived from USGS 1-arc second DEM [56].30 m
Table 2. DFA Predictive Accuracy. Gages with sufficient in situ data to determine intermittency status (perennial vs. intermittent channel) were used to construct the linear DFA for prediction of group membership. Known group membership was used to determine the predictive accuracy of the linear DFA. Jackknife resampling was further used to validate the predictive accuracy of the model.
Table 2. DFA Predictive Accuracy. Gages with sufficient in situ data to determine intermittency status (perennial vs. intermittent channel) were used to construct the linear DFA for prediction of group membership. Known group membership was used to determine the predictive accuracy of the linear DFA. Jackknife resampling was further used to validate the predictive accuracy of the model.
DFA Predictive AccuracyJackknife Predictive Accuracy
Individual GroupOverall
Perennial Channels82.3%79.5%74.3%
Intermittent Channels77.8%
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Davidson, L.J.; Milewski, A.M. Ephemeral Channel Expansion: Predicting Shifts Toward Intermittency in Vulnerable Streams Across Semi-Arid CONUS. Water 2025, 17, 3445. https://doi.org/10.3390/w17233445

AMA Style

Davidson LJ, Milewski AM. Ephemeral Channel Expansion: Predicting Shifts Toward Intermittency in Vulnerable Streams Across Semi-Arid CONUS. Water. 2025; 17(23):3445. https://doi.org/10.3390/w17233445

Chicago/Turabian Style

Davidson, Lea J., and Adam M. Milewski. 2025. "Ephemeral Channel Expansion: Predicting Shifts Toward Intermittency in Vulnerable Streams Across Semi-Arid CONUS" Water 17, no. 23: 3445. https://doi.org/10.3390/w17233445

APA Style

Davidson, L. J., & Milewski, A. M. (2025). Ephemeral Channel Expansion: Predicting Shifts Toward Intermittency in Vulnerable Streams Across Semi-Arid CONUS. Water, 17(23), 3445. https://doi.org/10.3390/w17233445

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop