Wilderness stream crossings used by mules, horses, and hikers are localized disturbances that may affect habitat immediately downstream, but the potential influence of fords on streams has received little investigation, particularly in terms of possible effects on fauna. Our overall null hypothesis was absence of below-above differences for either benthic macroinvertebrate assemblages or habitat characteristics at such fords. We further sought to determine (1) whether any such differences were present prior to annual use, suggesting longer-term effects, and (2) whether differences were present in late season, after annual use. We examined macroinvertebrates and habitat immediately below and above 2 fords crossing subalpine streams in Yosemite National Park in the Sierra Nevada (California, USA) in earlyand late season and over 2 years. There were both longer-term below-above differences, as well as differences thatbecame apparent in late season, both of which were indicative of below-ford effects. Below fords there was evidence,either as main effects or interactions, of higher silt, sand, and gravel cover; a thicker periphyton layer; a greater Hilsenhoffbiotic index; a higher proportion of tolerant taxa; higher chironomid midge and total densities; and greater speciesrichness, largely a function of chironomid richness. There was also a lower expected number of species, a smaller proportionof sensitive taxa and predators, and lower densities of some sensitive Ephemeroptera (mayflies) and Plecoptera (stoneflies) below fords. Both hikers and stock may contribute to the apparent effects, but management interventions targeting stock may be particularly achievable. Among other approaches, simply halting stock strings briefly before reaching fords should reduce the volume of urine and feces directly entering streams, and handlers can expedite crossings if watering is not necessary.

We investigated the potential magnitude and duration of forest evapotranspiration (ET) decreases resulting from forest-thinning treatments and wildfire in west-slope watersheds of the Sierra Nevada range in California, USA, using a robust empirical relation between Landsat-derived mean-annual normalized difference vegetation index (NDVI) and ET measured at flux towers. Among forest treatments, the minimum observed NDVI change required to produce a significant departure from control plots with NDVI of about 0.70 was −0.09 units, corresponding to a basal-area reduction of 29.1 m2/ha (45% reduction) and equivalent to an estimated ET reduction of 153 mm/year (21% change; approximate mean annual precipitation = 1,000 mm). Intensive thinning in highly productive forests that approached prefire-exclusion densities reduced basal area by 40–50%, generating estimated ET reductions of 153–218 mm/year (21–27% change) over 5 years following treatment. Low-intensity underburn treatments resulted in no significant change in ET. Examining the cumulative impact of wildfires on ET between 1990 and 2008, we found that the lower and wetter American River basin (5,310 km2) generated more than twice the ET reduction per unit area than those in the higher and drier Kings River basin (4,790 km2), corresponding to greater water and energy limitations in the latter and greater fire severity in the former. A rough extrapolation of these results to the entire American River watershed suggests that ET reductions due to forest thinning by wildfire could approach 10% of full natural flows for dry years and 5% over all years.

We assessed the response of densely forested watersheds with little apparent annual water limitation to forest disturbance and climate variability, by studying how past wildfires changed forest evapotranspiration, and what past evapotranspiration patterns imply for the availability of subsurface water storage for drought resistance. We determined annual spatial patterns of evapotranspiration using a top-down statistical model, correlating measured annual evapotranspiration from eddycovariance towers across California with NDVI (Normalized Difference Vegetation Index) measured by satellite, and with annual precipitation. The study area was the Yuba and American River watersheds, two densely forested watersheds in the northern Sierra Nevada. Wildfires in the 1985-2015 period resulted in significant post-fire reductions in evapotranspiration for at least 5 years, and in some cases for more than 20 years. The levels of biomass removed in medium-intensity fires (25- 75% basal area loss), similar to magnitudes expected from forest treatments for fuels reduction and forest health, reduced evapotranspiration by as much 150-200 mm yr-1 for the first 5 years. Rates of recovery in post-wildfire evapotranspiration confirm the need for follow-up forest treatments at intervals of 5-20 years to sustain lower evapotranspiration, depending on local landscape attributes and interannual climate. Using the metric of cumulative precipitation minus evapotranspiration (P-ET) during multi-year dry periods, we found that forests in the study area showed little evidence of moisture stress during the 1985-2018 period of our analysis, owing to relatively small reliance on interannual subsurface water storage to meet dry-year evapotranspiration needs of vegetation. However, more-severe or sustained drought periods will push some lower-elevation forests in the area studied toward the cumulative P-ET thresholds previously associated with widespread forest mortality in the southern Sierra Nevada.

Abstract: In order to investigate snowpack sensitivity to temperature increases and end‐member atmospheric moisture conditions, we applied a well‐constrained energy‐ and mass‐balance snow model across the full elevation range of seasonal snowpack using forcing data from recent wet and dry years. Humidity scenarios examined were constant relative humidity (high) and constant vapor pressure between storms (low). With minimum calibration, model results captured the observed magnitude and timing of snowmelt. April 1 snow water equivalent (SWE) losses of 38%, 73%, and 90% with temperature increases of 2, 4, and 6°C in a dry year centered on areas of greatest SWE accumulation. Each 2°C increment of warming also resulted in seasonal snowline moving upslope by 300 m. The zone of maximum melt was compressed upward 100–500 m with 6°C warming, with the range reflecting differences in basin hypsometry. Melt contribution by elevations below 2,000 m disappeared with 4°C warming. The constant‐relative‐humidity scenario resulted in 0–100 mm less snowpack in late spring vs. the constant‐vapor‐pressure scenario in a wet year, a difference driven by increased thermal radiation (+1.2 W/m2) and turbulent energy fluxes (+1.2 W/m2) to the snowpack for the constant‐relative‐humidity case. Loss of snowpack storage and potential increases in forest evapotranspiration due to warming will result in a substantial shift in forest water balance and present major challenges to land management in this mountainous region.

Watershed managers require accurate, high-spatial-resolution evapotranspiration (ET) data to evaluate forest susceptibility to drought or catastrophic wildfire, and to determine opportunities for enhancing streamflow or forest resilience under climate warming. We evaluate an easily calculated product by using annual gridded precipitation (P) and measured discharge (Q), together with a gridded ET product developed from ET and P measured at flux towers plus Landsat NDVI (normalized difference vegetation index) to evaluate uncertainties in water balances across 52 watersheds with stream-gauge measurements in the Central Sierra Nevada. Watershed areas ranged from 5 to 4823 km2, and the study-area elevation range was 52–3302 m. Study-area P, ET, and Q averaged 1263, 634, and 573 mm yr–1 respectively, with precipitation at higher elevations up to five times that at lower elevations. We assessed uncertainty in water-balance components by applying a multiplier to P or Q values across the period of record for each watershed to align annual P-ET and Q values, resulting in average P-ET-Q = 0. Most year-to-year values of annual change in storage (ΔS), calculated as P-ET-Q for watersheds with well-constrained water balances, were within about + 300 mm. Across the study area we found that for each of 37 watersheds, applying a constant multiplier to either annual P or Q resulted in well-constrained water balances (average annual P-ET-Q = 0). Multiplicative adjustment of ET values for each watershed did not improve average water balances over the period of record, and would result in inconsistent values across adjacent and nested watersheds. For a given watershed, ET was relatively constant from year to year, with precipitation variability driving both interannual and spatial variability in runoff. These findings highlight the importance of evapotranspiration as a central metric of water-balance change and variability, and the strength of using high-confidence spatial- evapotranspiration estimates to diagnose uncertainties in annual water balances, and the components contributing to those uncertainties.

Abstract. We present hourly climate data to force land surface process models and assessments over the Merced and Tuolumne watersheds in the Sierra Nevada, California, for the water year 2010–2014 period. Climate data (38 stations) includes temperature and humidity (23), precipitation (13), solar radiation (8), and wind speed and direction (8) spanning an elevation range of 333 to 2987 m. Each data set contains raw data as obtained from the source (level 0), data that are serially continuous with noise and non-physical points removed (level 1), and, where possible, data that are gap-filled using linear interpolation or regression with a nearby station record (level 2). All stations chosen for this data set were known or documented to be regularly maintained and components checked and calibrated during the period. Additional time-series data included are available snow water equivalent records from automated stations (8) and manual snow courses (22), as well as distributed snow-depth and co-located soil-moisture measurements (2–6) from four locations spanning the rain-to-snow transition zone in the centre of the domain. Spatial data layers pertinent to snowpack modelling in this data set are basin polygons and 100-m resolution rasters of elevation, vegetation type, forest basal area, tree height, and forest canopy cover, transmissivity, and extinction coefficient. All data are available from online data repositories (https://doi.org/10.6071/M3FH3D).

We present hourly climate data to force land surface process models and assessments over the Merced and Tuolumne watersheds in the Sierra Nevada, California, for the water year 2010-2014 period. Climate data (38 stations) include temperature and humidity (23), precipitation (13), solar radiation (8), and wind speed and direction (8), spanning an elevation range of 333 to 2987 m. Each data set contains raw data as obtained from the source (Level 0), data that are serially continuous with noise and nonphysical points removed (Level 1), and, where possible, data that are gap filled using linear interpolation or regression with a nearby station record (Level 2). All stations chosen for this data set were known or documented to be regularly maintained and components checked and calibrated during the period. Additional time-series data included are available snow water equivalent records from automated stations (8) and manual snow courses (22), as well as distributed snow depth and co-located soil moisture measurements (2-6) from four locations spanning the rain-snow transition zone in the center of the domain. Spatial data layers pertinent to snowpack modeling in this data set are basin polygons and 100 m resolution rasters of elevation, vegetation type, forest canopy cover, tree height, transmissivity, and extinction coefficient. All data are available from online data repositories (https://doi.org/10.6071/M3FH3D).