Naomi E. Detenbeck

High-resolution temperature models are needed to describe and predict the thermal regime of streams/rivers under current and future management. Statistical models (e.g., spatial stream network models – SSN) can cover large extents at fine resolutions. Mechanistic models (e.g., Soil and Water Assessment Tool – SWAT) offer refined temporal resolution and process-based drivers in predictions, but are time consuming to parameterize. Hybrid modeling (e.g., hybrid SSN-SWAT) approaches can leverage benefits of both types to improve temporal robustness of thermal regime model predictions across entire networks. In the Upper Rogue River Basin, OR, we compared performance of a NorWeST August mean temperature (AMT) SSN model with a hybrid SSN-SWAT AMT (Hybrid AMT) model. We extended the hybrid SSN-SWAT approach to the 7-day average daily maximum temperature metric (Hybrid 7DADM). The Hybrid AMT and Hybrid 7DADM models performed well, with similar model diagnostics between the Hybrid AMT and NorWeST AMT models, and with other SSN and mechanistic temperature models (used for temperature total maximum daily loads – TMDLs). The hybrid SSN-SWAT approach was able to incorporate the effect of dams on downstream hydrology and temperature regimes. The best-fitting hybrid SSN models incorporated upstream shade effects spanning distances of 1-5 kilometers or 1-3 hours travel time, highlighting the need to maintain upstream riparian zones to protect downstream thermal regimes.

Forest practices can exacerbate or moderate risks to coldwater habitat (CWH) related to climate variability, wildfire, and forest disease. We applied a high-resolution hybrid mechanistic/statistical stream network (SSN) model to evaluate effects of riparian management practices and future climate in the Upper Rogue, OR basin on flow and temperature. We compared recent (1990 – 2018) and 2040’s climate scenarios for temperature and precipitation, considering both average, minimum, and maximum based on the Palmer Drought Severity Index. Riparian management conditions included historical reference, current vegetation (CV), and vegetation managed via the Oregon Forest Practices (FPA) Act which prescribes minimum forest buffer widths of 100-110 ft for fish-bearing (F) and 35 ft for non fish-bearing (NF) streams. Under the 2040s scenario, flow will be lost by the end of August in 79% of F reaches and 93% of NF reaches in a dry year. Interannual variability of CWH extent will increase during the 2040s. In cool/wet years, core F CWH (< 16 0C) extent will decline by ~ 10 % while habitat > 18 0C will increase under the FPA scenario as compared to the current vegetation regime. For 2040s flowing streams under mean climate, the extent of NF reaches with <15 °C 7DADM will be reduced under the FPA to 2.2% of reaches as compared to current vegetation (4.5 % of reaches). Thus, reduced protections prescribed for NF streams will impact downstream CWH as well as headwater NF habitat.

Cyanobacterial blooms have been increasingly detected in estuaries and freshwater tidal rivers. To enhance detailed monitoring, an efficient approach to detecting algal blooms through remote sensing is needed to focus more detailed monitoring focused on cyanobacteria. In this study, we compared different remote sensing processing methods to determine an efficient approach to mapping chlorophyll-a. Using a subset of paired chlorophyll-a observations with Sentinel-2 imagery (2015–2022), with sites located in the Chesapeake Bay and Indian River selected along gradients of salinity, turbidity, and trophic status, we compared the combined performance of two different atmospheric processing methods (Acolite, Polymer) and a suite of empirical (band ratio, spectral shape indices) and machine learning algorithms for chlorophyll-a prediction. Acolite outperformed Polymer, resulting in 176 observation points, compared to 106 observation points from Polymer, and a greater range in chlorophyll-a values (0–74 μg/L from Acolite compared to 0–36 μg/L from Polymer), although Polymer showed more responsiveness at lower chlorophyll-a levels. Two algorithms performed best in predicting chlorophyll-a, as well as trophic state and HABs risk classes: the machine learning mixture density network (MDN) approach and the one band-ratio approach (Mishra).

This study was developed to identify the optimal (most cost-effective) strategies to reduce sediment and phosphorus loadings in the Upper Soldier Creek, Kansas, USA, watershed using the Watershed Management Optimization Support Tool (WMOST) suite of programs. Under average precipitation, loading targets for upland total phosphorus (TP) could be met with use of grassed swales for treating urban area runoff and of contouring for agricultural runoff. For a wet year, the same target could be met, but with use of a sand filter with underdrain for the urban runoff. Both annual and daily TP loading targets from Total Maximum Daily Loads (TMDLs) were exceeded in simulations of best management practice (BMP) solutions for 14 alternative future climate scenarios. We expanded the set of BMPs to include stream bank stabilization (physical plus riparian restoration) and two-stage channel designs, but upland loading targets could not be met for either TP or total suspended solids (TSS) under any precipitation conditions. An optimization scenario that simulated the routing of flows in excess of those treated by the upland BMPs to an off-channel treatment wetland allowed TMDLs to be met for an average precipitation year. WMOST can optimize cost-effectiveness of BMPs across multiple scales and climate scenarios.

Researchers and environmental managers need big datasets spanning long time periods to accurately assess current and historical water quality conditions in fresh and estuarine waters. Using remote sensing data, we can survey many water bodies simultaneously and evaluate water quality conditions with greater frequency. The combination of existing and historical water quality data with remote sensing imagery into a unified database allows researchers to improve remote sensing algorithms and improves understanding of mechanisms causing blooms. We report on the development of a water quality database “EstuarySAT” which combines data from the Sentinel-2 multi-spectral instrument (MSI) remote sensing platform and water quality data throughout the coastal USA. EstuarySAT builds upon an existing database and set of methods developed by the creators of AquaSat, whose region of interest is primarily larger freshwater lakes in the USA. Following the same basic methods, EstuarySAT utilizes open-source tools: R v. 3.24+ (statistical software), Python (dynamic programming environment), and Google Earth Engine (GEE) to develop a combined water quality data and remote sensing imagery database (EstuarySAT) for smaller coastal estuarine and freshwater tidal riverine systems. EstuarySAT fills a data gap that exists between freshwater and estuarine water bodies. We are able to evaluate smaller systems due to the higher spatial resolution of Sentinel-2 (10 m pixel image resolution) vs. the Landsat platform used by AquaSat (30 m pixel resolution). Sentinel-2 also has a more frequent revisit (overpass) schedule of every 5 to 10 days vs. Landsat 7 which is every 17 days. EstuarySAT incorporates publicly available water quality data from 23 individual water quality data sources spanning 1984–2021 and spatially matches them with Sentinel-2 imagery from 2015–2021. EstuarySAT currently contains 299,851 matched observations distributed across the coastal USA. EstuarySAT’s primary focus is on collecting chlorophyll data; however, it also contains other ancillary water quality data, including temperature, salinity, pH, dissolved oxygen, dissolved organic carbon, and turbidity (where available). As compared to other ocean color databases used for developing predictive chlorophyll algorithms, this coastal database contains spectral profiles more typical of CDOM-dominated systems. This database can assist researchers and managers in evaluating algal bloom causes and predicting the occurrence of future blooms.

Under the United States Clean Water Act, states are required to periodically assess state waters to determine compliance with water quality criteria (including temperature) and then to develop total maximum daily loads (TMDLs) for impaired waters as necessary to bring them into compliance. We compared the performance of mechanistic stream temperature models (HeatSource, QUAL2K, and QUAL2Kw) applied to the mainstem of three TMDL watersheds (Middle Fork John Day, OR; Wind River, WA; South Fork Nooksack, WA) with that of spatial stream network (SSN) models applied to the full watersheds and used these to evaluate the potential effectiveness of restoration strategies. SSN models performed well with slightly lesser accuracy (RMSE = 0.47–0.87) for mainstem predictions than mechanistic models (RMSE = 0.4) but provided additional benefits to inform management, including information on spatial and temporal heterogeneity of restoration effectiveness throughout the watershed. Of the four scenarios considered (restoration of riparian zones to potential natural vegetation, channel narrowing, increasing flow by restricting irrigation withdrawals, and combined applications), riparian zone restoration was consistently the most effective in reducing temperatures at the outlet, mainstem, and throughout the watersheds. Predicted restoration effectiveness for thermal regimes varied significantly both within and among watersheds. A focus on water quality criteria exceedance only at the watershed outlet or along the mainstem reach can obscure knowledge of restoration potential for fish habitat in tributaries and headwaters, potential for creation of thermal refuge areas along the mainstem critical for maintaining migration corridors, and thermal regime heterogeneity across space and time.

Many cold water‐dependent aquatic organisms are experiencing habitat and population declines from increasing water temperatures. Identifying mechanisms which drive local and regional stream thermal regimes facilitates restoration at ecologically relevant scales. Stream temperatures vary spatially and temporally both within and among river basins. We developed a modeling process to identify statistical relationships between drivers of stream temperature and covariates representing landscape, climate, and management‐related processes. The modeling process was tested in three study areas of the Pacific Northwest United States during the growing season (May [start], August [warmest], September [end]). Across all months and study systems, covariates with the highest relative importance represented the physical landscape (elevation [1st], catchment area [3rd], main channel slope [5th]) and climate covariates (mean monthly air temperature [2nd] and discharge [4th]). Two management covariates (groundwater use [6th] and riparian shade [7th]) also had high relative importance. Across the growing season (for all basins), local reach slope had high relative importance in May, but transitioned to a regional main channel slope covariate in August and September. This modeling process identified regionally similar and locally unique relationships among drivers of stream temperature. High relative importance of management‐related covariates suggested potential restoration actions for each system.

Excess loading of nitrogen and phosphorus to river networks causes environmental harm, but reducing loads from large river basins is difficult and expensive. We developed a new tool, the River Basin Export Reduction Optimization Support Tool (RBEROST) to identify the least‐cost combinations of management practices that will reduce nutrient loading to target levels in downstream and mid‐network waterbodies. We demonstrate the utility of the tool in a case study in the Upper Connecticut River Basin in New England, USA. The total project cost of optimized lowest‐cost plans ranged from $18.0 million to $41.0 million per year over 15 years depending on user specifications. Plans include both point source and non‐point source management practices, and most costs are associated with urban stormwater practices. Adding a 2% margin of safety to loading targets improved the estimated probability of success from 37.5% to 99%. The large spatial scale of RBEROST, and the consideration of both point and non‐point source contributions of nutrients, make it well suited as an initial screening tool in watershed planning.

Background: Taxa Indicator Threshold ANalysis (TITAN) was developed to identify thresholds along environmental gradients where rapid changes in taxa frequency and relative abundance are observed. TITAN determines separate change-points for increasing and decreasing taxa in aggregate, as well as change-points for individual taxa, with associated confidence intervals generated using bootstrapping. However, if TITAN is applied to different classes of observations, additional analyses besides using non-overlapping confidence intervals are needed to establish whether change-points differ between treatments or groups because non-overlapping confidence intervals can indicate significant differences but overlapping confidence intervals do not necessarily mean the null hypothesis cannot be rejected. Methods: To address this, we present a new R package, pTITAN2, which is an extension to the existing TITAN2 package. The pTITAN2 package was developed to enable comparisons of TITAN output between treatments by permutating the observed data between treatments and rerunning TITAN on the permuted data. Results: The pTITAN2 package includes two functions, occurrences and permute. The occurrences function selects the taxonomic codes to be used in a TITAN run while maintaining the most taxonomic details. The permute function is then used to create a list of permuted sets of taxa and environmental gradients. TITAN is then run again on the permuted data and p-value test can be calculated using the observed and permuted TITAN output to test for statistical differences between treatment effects. Conclusions: The package pTITAN2 is an extension of the existing TITAN2 package and enables users to perform the appropriate statistical tests and determine statistical differences without using overlapping confidence intervals.