Water-supply outlook: February bad, March good?

At the beginning of February, we reported a reasonable chance of ending up with near-average snowpack by April 1. How are we doing a month later?

Snowpack

As expected, precipitation during the month of February was below average, but it turned out to be even lower than we had anticipated. As a result, snow-water equivalent (SWE) at sites in the upper Henry's Fork region is much lower with respect to average now than it was one month ago. Last week, the forecast for March was dry, and using that prediction, we arrived at a projection for April 1 SWE that isn't nearly as close to average as the one we provided last month. The following table compares SWE at four snow-survey sites in and near Island Park to historic median and mean values, as measured in early and late February, and as we project it to be on April 1.

Table of SWE values at snow survey sites near Island Park.

The table shows that as a percentage of both long-term median and mean values, SWE has declined since February 1 and is expected to continue to decline. We project that SWE on April 1 will be around 80-90% of long-term means at all four snow-survey sites in the Island Park vicinity. We include both medians and means in this table to illustrate that the median is smaller than the mean, so that percentages of median provide an overly optimistic view of water supply relative to the mean. Thus, when you look at the daily SWE values on the Natural Resources Conservation Service web site, keep in mind that the percentages are relative to the 30-year median, which is a little smaller than the long-term mean.

Inflow to Island Park Reservoir

As I've discussed in previous blogs, the two snow-survey sites that have the greatest statistical power to predict spring-time (April-June) inflow to Island Park Reservoir are the Crab Creek and Black Bear sites. Crab Creek is located in the Centennial Mountains, and Black Bear is on the western edge of the Yellowstone Plateau, between Big Springs and West Yellowstone.

Based on our projected April 1 SWE at these two sites, and on observed baseflows in the upper Henrys Fork this winter, we are now projecting that April-to-June inflow to Island Park Reservoir will be only 70% of average. That's better than last year's 56% of average but still not great. The graph below shows our projection, in comparison to all other water years since 1979. The x-axis is winter baseflow, which is the flow provided by groundwater and is reflective of cumulative precipitation over the last three years.

Graph of spring-time inflow to Island Park Reservoir versus previous winter baseflow.

The primary reason why our prediction for the spring of 2016 is so low is because baseflows are near record lows due to the past three years of drought. This is evidenced by how far to the left our predicted value (the red diamond in the graph) is on the graph. You will also notice that the red diamond falls a little bit below the prediction curve. That is because the curve in this graph applies to years with average snowpack, and as indicated in the table above, we expect snowpack to be a little below average.

Outflow from Island Park Reservoir

So, with a predicted April-to-June inflow that is going to average only around 600 cfs, what does that mean for outflow from Island Park Reservoir? Our projections for inflow, outflow, and reservoir volume are shown on the next graph, which also shows average values, for comparison.

Graph of inflow to, outflow from, and volume in Island Park Reservoir, as predicted for spring 2016.

The left panel of the graph shows that inflow and outflow have been much lower than average all winter and are expected to remain below average throughout the spring. The right panel shows that at current inflow and outflow, the reservoir will not meet our April 1 target of 127,000 acre-feet (94% full) until the middle of April. However, we do not anticipate the need to decrease outflow to achieve the target any sooner than mid-April.

Once the 127,000 target is achieved, outflow will be increased to roughly match inflow and allow the remaining 6% of reservoir capacity to fill slowly throughout late April and remain constant near its full capacity throughout May. With careful planning and management, coordination among members of the Island Park Drought Management Planning Committee, and a litte luck, outflow from the reservoir will be very close to inflow--that is, the river's natural flow--during late May and June. That's the good news--outflow will simply reflect the watershed's natural hydrology during the late spring. The bad news is that this natural flow is still well below average because of drought, so flows will be around 500 cfs in the middle of June, even with a full reservoir.

What could change?

First, snowfall in March could be quite a bit different than the slightly-below-average value we used to predict the values shown in the table above. As an example, consider the Black Bear site. As of today, March 3, SWE is 24.5 inches. The minimum snowfall between March 3 and April 1 recorded at that site over the past 44 years is 0.9 inches, and the maximum is 15.5 inches. Using these as extreme cases, it is nearly certain that April 1 SWE at that site will end up somewhere between 25.4 inches (66% of average) and 40 inches (104% of average). Our projection is just a little below the middle of these two extremes. For comparison, April 1 SWE at Black Bear in 2015 was just 25 inches. So, the good news is that even if March is as dry as it has ever been, we'll end up better than last year. The bad news is that it will take close to the wettest March on record to put the April 1 SWE at or above average. The latest medium-range forecast is calling for March precipitation to be near average, so our prediction of around 30 inches (around 80% of average) is now pretty likely to happen.

Second, spring-time precipitation--whether it be rain or snow--is a big unknown. Although inflow to Island Park Reservoir is primarily dependent on snowmelt and on baseflow, spring-time precipitation does have a small effect on inflow. If the spring is much drier than average, our prediction will prove to be a little optimistic, and if the spring is wetter than average, then streamflow will be a little higher than we predicted. The current long-term forecast is for average precipitation during the spring, so most likely, our prediction will be pretty accurate.

Third, no matter what the water supply looks like, management of Island Park Reservoir is highly dependent on early-season irrigation demand. Last year was the best example of this on record. In 2015, warm and dry weather during March and April followed a warm, dry winter that produced very little snow cover in the valleys. The result was very early planting and record high irrigation demand on April 1, the first day of the legal irrigation season. That high demand, in combination with record low natural streamflows in Fall River and the upper Henry's Fork, required delivery of irrigation water from Island Park Reservoir beginning in early June. Although it is still possible that something like this could happen in 2016, the chances are getting smaller every day. There is still good snow cover in the valley from Rexburg northward, and that snow will not completely melt for at least another two to four weeks, depending on elevation. The latest weather forecasts are calling for cool temperatures next week, only slightly above-average temperatures over the next two weeks, and above-average precipitation over that same time period. Lastly, even with below-average snowpack, natural streamflows are expected to be between 70% and 90% of average in the upper Snake River basin. While still below average, this is much better than streamflows over most of the basin in 2015. Thus, at this point, we do not anticipate the need for early irrigation delivery from Island Park Reservoir.