We have developed a conceptual model of the contribution of soil moisture to spring runoff and now have seven years of soil moisture monitoring to support the concepts. The drought conditions have provided opportunities to refine the simulation tools for accuracy in calculating the water balance and runoff, including revisions in soil depth to better reflect rooting zones. Preliminary analyses have been done for locations in the Russian River, but Yosemite locations have not yet been accessed for assessment of the drought years 2015 and 2016. We will extend these analyses to all available data (Hydrometeorological Testbed, HMT; National Resource Conservation Service, NRCS; Critical Zone Observatories (CZO); Scripps Institution of Oceanography) to extrapolate across the state. Prior to completing the documentation, we will incorporate these validation efforts and improve simulation tools for spring forecasting.

• Collect soil-moisture data at locations across the state to provide time-series data with which to calibrate the revised dry-out capability in the BCM as a function of varying climatic water deficit (CWD).
• Conduct sensitivity analyses of soil moisture below wilting point for 2013-2015 with changes in CWD. Develop drying curves associated with soil types or properties statewide.
• Develop approach to improve estimates of rooting depth, incorporating new soil layer, and using estimates of actual evapotranspiration (such as the METRIC, Sebal, or Normalized Difference Vegetation Index [NDVI] data).
• Validate runoff estimates using improved precipitation data and model revisions with reservoir inflows and streamflow data for selected locations.

An approach has been developed to optimize monitoring locations that have been tested in the Lake Mendocino basin with the intent of distribution throughout the Russian River basin. Similarly for Hetch Hetchy, using the updated soil properties maps and model revisions, and relying on the Dana Meadows and Gin Flat soil potential/moisture data as forecasting tools, we will determine conditions under which these data can be used to constrain forecasting simulations, evaluate uncertainties, and identify additional locations to provide improved forecasts in conjunction with ASO data for the upper Tuolumne (Hetch Hetchy) basin in the San Joaquin river basin. This approach would then be applied statewide to optimize soil-moisture monitoring strategies for early warning purposes.

• Analyze conditions in the Hetch Hetchy basin using soil-moisture data, test utility for forecasting, and identify additional soil-moisture locations that would significantly improve forecasts.
• Identify and prepare data layers similar to those used in the Lake Mendocino basin exercise for the State of California, considering the varying differences in sensitivities to extremes across the state (both drought and flood are necessary for purposes of early warning), and using omission overlays for urban, agriculture, and other unrealistic locations to monitor soil moisture. Highlight reservoirs and other locations of interest for forecasting runoff and CWD, and include existing infrastructure, preserves, and ongoing soil-moisture data collection platforms.
• Perform exercise to locate about 200 soil moisture stations using weighted average binning (discuss with DWR to determine the optimum number and locations of interest).
• If data allow, we will consider how and where to monitor springs in conjunction with wells to monitor groundwater conditions.

Having the dry soil conditions included in the revised modeling tools (item 2), we will apply monthly revisions to the daily BCM, assimilate JPL/SWE data for the entire Sierra Nevada (and the Pit River basin in the Sacramento River basin in northeastern California if the JPL data allows), and recalibrate daily models to accumulate SWE and accurately reflect the water balance for spring snow melt. Results for 2000-2014 will be compared to the 8-station and 5-station indices and measured streamflow throughout the Sierra. Building on previous work, we'll tie the SWE in with antecedent hydrologic conditions and provide an improved forecasting tool for spring runoff volumes. We will further calibrate the solar radiation function using the JPL and ASO data available for the Tuolumne basin, which would be included in the final documentation.

• Collect all data and revise daily BCM.
• Provide calibration results to illustrate improved snow accumulation and melt capability for Sierra Nevada basins and how it matches JPL snow covered area and snow-water equivalent. Identify additional issues with simulations of snow, indicate the need to have JPL include the Pit River basin, and assess the need to include eastern Sierras in the daily modeling.
• Evaluate differences between using a solar flag or not, and if it improves estimates in the upper Tuolumne.
• Develop map of spatially distributed snow parameters.

Develop a template for a California statewide webpage with monthly updated situational status maps that will inform emergency response and planning. It needs to be monthly, reproducible, validated or correlated to local information, and not rely on current conditions of the managed system. The webpage may be produced in conjunction with the Western Regional Climate Center administered by NOAA in Reno at the Desert Research Institute that hosts the Southwest Climate and ENvironmental Information Collaborative (SCENIC), or on the USGS California Water Science Center home page. This determination hasn't been made yet.

• Determine needs for ER Intel iteratively with DWR.
• Design template to include maps representing status of water supply, landscape drought and natural demand, wildfire risk and potentially reservoirs and groundwater conditions. Determine necessary explanations, validation information, associated information and links.
• Develop approach to represent cumulative climatic water deficit over time on a grid-cell basis.
• Collect and analyze forage-condition data to illustrate relevance with CWD.
• Develop validation exercises, including both water use and landscape health/stress:
• Vegetation drought response index and CWD.
• Cumulative Normalized Difference Vegetation Index (NDVI) or Enhanced Vegetation Index (EVI) and CWD.
• Changes in groundwater recharge and corresponding groundwater levels.
• Discussion of CWD used to illustrate increases in population and agricultural water demand as a result of warming climates.
• Demonstrate ability to use BCM simulated recharge plus runoff to reflect water supply using measured data, comparisons with reservoir inflows and soil-water storage, etc.
• Develop framework for simple spot-check validation (perhaps through provided links to measured soil moisture and SWE station indices).
• Investigate the possibility or usability of a drought severity binning approach. Possible example is using CWD and BCM estimated recharge plus runoff in percentiles according to historical conditions; as an example, compare with historical archives of the U.S. Drought Monitor.
• Because soil-water content is correlated to forage condition statewide, collect foragecondition data from Mel George (University of California, Davis emeritus) to validate against BCM indices. Link to soil-moisture data for purposes of early warning.