skip to main content

Title: High‐resolution temperature modeling of stream reconstruction alternatives

Streams are complex where biology, hydrology, and atmospheric processes are all important. Because quantifying and modeling of these systems can be challenging, many teams go directly to prescribed restoration treatments and principles. Restoration on the Middle Fork of the John Day River in Oregon, USA, shows how a project that was designed according to widely accepted restoration principles may lead to outcomes contrary to one of the project's stated goals: reducing peak temperatures for endangered salmonids on the site. This study employed the most sophisticated equipment available for stream temperature monitoring, including approximately 1 million independent hourly measurements in the 2‐week period considered. These data were collected along the river channel with fiber optic–distributed temperature sensing and were used to quantify thermal dynamics. These observations were paired with a physically based stream temperature model which was then employed to predict temperature change from design alternatives. Restored‐reach impact on peak temperature was directly correlated with the air–water interfacial area and the percentage of effective shade (R2 > 0.99). The increase in air–water area of the proposed design was predicted to increase daytime stream temperature by as much as 0.5°C upon completion of the work. Shade from riparian vegetation was found to potentially mitigate stream temperature increases, though only after decades of growth. A moderately dense canopy of 5 m tall trees blocking 17% of daily shortwave solar radiation is predicted to mitigate predicted temperature increases over the 1,800 m reach but also increases nighttime temperatures due to blocking of long‐wave radiation. These outcomes may not be intuitive to restoration practitioners and show how quantitative analysis can benefit the design of a project. This is significant in an area where riparian vegetation has been difficult to reestablish. Without quantitative analysis, restoration efforts can lead to outcomes opposite to stated goals and may be costly and disruptive interventions to fragile stream systems.

more » « less
Award ID(s):
Author(s) / Creator(s):
Publisher / Repository:
Wiley Blackwell (John Wiley & Sons)
Date Published:
Journal Name:
River Research and Applications
Page Range / eLocation ID:
p. 931-942
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    Rivers are dynamic, complex integrators of their environment, which makes verification of the beneficial outcomes of restoration challenging. Thermal regime is central to habitat suitability and is often a focus in planning and evaluating the impact of restoration and climate resilience. Among these concerns, high summer stream temperature has frequently been identified as a limiting factor for salmon, steelhead, and trout. Our objective was to demonstrate the utility of combining high resolution thermal observation and modelling to evaluate restoration designed to mitigate stream thermal processes. This was demonstrated on the Middle Fork of the John Day River which is a critically impacted salmonid fishery in northeast Oregon, USA. We employed distributed temperature sensing and energy‐balance modelling to define the thermal regime. Restoration was predicted to result in a 0.7°C reduction of peak daily stream temperatures while increasing night temperatures by 0.9°C. This combined modelling and monitoring approach suggests that the 2012 restoration offered relief for native fish species stressed by excessive stream temperatures. This powerful combination of technology can be used in many projects to make optimal use of restoration investments to achieve durable and quantifiable improvements in habitat.

    more » « less
  2. Abstract

    How much stream temperatures increase within riparian canopy openings and whether stream temperatures cool downstream of these openings both have important policy implications. Past studies of stream cooling downstream of riparian openings have found mixed results including rapid, slow, and no cooling. We collected longitudinal profiles of stream temperatures above, within, and below riparian forest openings along stream segments within otherwise forested riparian conditions to evaluate how sensitivity of stream temperatures to riparian conditions varied across landscape factors. We conducted these temperature surveys across openings in 12 wadeable streams within and near the Upper Little Tennessee River Basin in western North Carolina and northeastern Georgia. Basin areas ranged from 74 to 6,913 ha, and bankfull channel widths varied from 3.4 to 16.4 m. Stream temperatures were collected every 15 min using HOBO® data loggers for 2 weeks in each stream, repeated later in summer in some streams. Reference temperatures were highest in stream reaches at low elevations and with large drainage areas. Stream temperature increases in the middle of riparian gaps were highest when streams drained small high‐elevation watersheds, and increases at the end of openings were highest when the opening length was large relative to watershed size. Downstream from openings, cooling rates were greatest in small, high‐elevation headwater streams and also increased with larger increases in canopy cover. Stream segments that warmed the most within openings also featured higher cooling rates downstream. The data show that stream temperature sensitivity to canopy change is highly dependent on network position and watershed size. A better understanding of stream temperature responses to riparian vegetation may be useful to land managers and landowners prioritizing riparian forest restoration.

    more » « less
  3. Given the widespread presence of non-native vegetation in urban and Mediterranean watersheds, it is important to evaluate how these sensitive ecosystems will respond to activities to manage and restore native vegetation conditions. This research focuses on Del Cerro, a tributary of the San Diego River in California, where non-native vegetation dominates the riparian zone, creating flooding and fire hazards. Field data were collected in 2018 to 2021 and consisted of water depth, streamflow, and stream temperature. Our data set also captured baseline conditions in the floodplain before and after the removal of burned non-native vegetation in November 2020. Observed changes in hydrologic and geomorphic conditions were used to parameterize and calibrate a two-dimensional hydraulic model to simulate urban floodplain hydraulics after vegetation removal. We utilized the U.S. Army Corps of Engineers’ Hydrologic Engineering Center River Assessment System (HEC-RAS) model to simulate the influence of canopy loss and vegetation disturbance and to assess the impacts of vegetation removal on stream restoration. We simulated streamflow, water depth, and flood extent for two scenarios: (1) 2019; pre-restoration where non-native vegetation dominated the riparian area, and (2) 2021; post-restoration following the removal of non-native vegetation and canopy. Flooding after restoration in 2021 was more frequent compared to 2019. We also observed similar flood extents and peak streamflow for storm events that accumulated half the amount of precipitation as pre-restoration conditions. Our results provide insight into the responses of small urban stream reaches to the removal of invasive vegetation and canopy cover. 
    more » « less
  4. Abstract

    Semi‐arid riparian woodlands face threats from increasing extractive water demand and climate change in dryland landscapes worldwide. Improved landscape‐scale understanding of riparian woodland water use (evapotranspiration, ET) and its sensitivity to climate variables is needed to strategically manage water resources, as well as to create successful ecosystem conservation and restoration plans for potential climate futures. In this work, we assess the spatial and temporal variability of Cottonwood (Populus fremontii)‐Willow (Salix gooddingii) riparian gallery woodland ET and its relationships to vegetation structure and climate variables for 80 km of the San Pedro River corridor in southeastern Arizona, USA, between 2014 and 2019. We use a novel combination of publicly available remote sensing, climate and hydrological datasets: cloud‐based Landsat thermal remote sensing data products for ET (Google Earth Engine EEFlux), Landsat multispectral imagery and field data‐based calibrations to vegetation structure (leaf‐area index, LAI), and open‐source climate and hydrological data. We show that at landscape scales, daily ET rates (6–10 mm day−1) and growing season ET totals (400–1,400 mm) matched rates of published field data, and modelled reach‐scale average LAI (0.80–1.70) matched lower ranges of published field data. Over 6 years, the spatial variability of total growing season ET (CV = 0.18) exceeded that of temporal variability (CV = 0.10), indicating the importance of reach‐scale vegetation and hydrological conditions for controlling ET dynamics. Responses of ET to climate differed between perennial and intermittent‐flow stream reaches. At perennial‐flow reaches, ET correlated significantly with temperature, whilst at intermittent‐flow sites ET correlated significantly with rainfall and stream discharge. Amongst reaches studied in detail, we found positive but differing logarithmic relationships between LAI and ET. By documenting patterns of high spatial variability of ET at basin scales, these results underscore the importance of accurately accounting for differences in woodland vegetation structure and hydrological conditions for assessing water‐use requirements. Results also suggest that the climate sensitivity of ET may be used as a remote indicator of subsurface water resources relative to vegetation demand, and an indicator for informing conservation management priorities.

    more » « less
  5. Abstract

    Dense understory thickets of the native evergreen shrubRhododendron maximumexpanded initially following elimination of American chestnut by the chestnut blight, and later in response to loss of the eastern hemlock due to hemlock woolly adelgid invasion. Rhododendron thickets often blanket streams and their riparian zones, creating cool, low‐light microclimates. To determine the effect of such understory thickets on summer stream temperatures, we removed riparian rhododendron understory on 300 m reaches of two southern Appalachian Mountain headwater streams, while leaving two 300 m reference reaches undisturbed. Overhead canopy was left intact in all four streams, but all streams were selected to have a significant component of dead or dying eastern hemlock in the overstory, creating time‐varying canopy gaps throughout the reach. We continuously monitored temperatures upstream, within and downstream of treatment and reference reaches. Temperatures were monitored in all four streams in the summer before treatments were imposed (2014), and for two summers following treatment (2015, 2016). Temperatures varied significantly across and within streams prior to treatment and across years for the reference streams. After rhododendron removal, increases in summer stream temperatures were observed at some locations within the treatment reaches, but these increases did not persist downstream and varied by watershed, sensor, and year. Significant increases in daily maxima in treatment reaches ranged from 0.9 to 2.6°C. Overhead canopy provided enough shade to prevent rhododendron removal from increasing summer temperatures to levels deleterious to native cold‐water fauna (average summer temperatures remained below 16°C), and local temperature effects were not persistent.

    more » « less