Water Planning Drenched in Uncertainty
With the sweat from 2011’s record-setting temperatures just now drying off and projections of a continued drought cycle remains for Central Texas, discussion about securing current and future water supplies is becoming increasingly urgent.
And not a moment too soon, as 2012 has had reports of at least one public water supply system in Texas failing and 13 others at risk of a similar fate within 180 days if the drought does not ease up. With so many external factors affecting the status of water supplies, there is a great deal of uncertainty involving their future.
To ensure future generations have sufficient water we need to first address the uncertainty in the method for developing water supply plans.
The Past Is Not the Future
Water management plans in Texas have a 50-year planning horizon and include predictions made for future water availability, as well as future water demands. The difference between the two — water need — is accounted for by proposing solutions via water management strategies, such as new reservoirs and well fields. But, there are improvements in the methodology for developing these plans that need to be implemented if we are to more effectively account for future conditions that may affect water supply and demand.
Water management plans in Texas are based off the assumption that the previous 50 years or so—including the devastating drought of the 1950s—are a fair model for the future. This is known as the “assumption of stationary,” which considers the magnitude and variability of the past a good representation of the future. It also assumes that there is not an underlying long-term trend.
The approach is purely deterministic rather than probabilistic; the predictions for water availability and demand are assumed to be known when in reality there is uncertainty on both sides of the equation. By taking the most devastating drought on record into consideration, it is assumed that a safety margin is in place, effectively a worst-case scenario, but the fact is that in the future we may experience droughts never seen before. And in fact the whole region may be experiencing a trend towards drier conditions, making our assumption of stationarity invalid and our safety margin insufficient.
Tree-ring studies and other paleohydrologic analyses already support the idea that next year’s drought could surpass the present drought of record. There are other significant areas of uncertainty in the predictions used for water planning beyond the status of the future climate calling into question the assumption that we’ll never see conditions as bad as we saw in the 1950s.
What Will You Need Tomorrow?
The current model for water planning is based on projected demand for water. But, uncertainties in that demand could cause plans to fall short. Demand projections are typically calculated separately for specific uses: municipal, industrial, irrigation, livestock and steam-electric.
Each type of demand is dependent on a number of factors. Population growth and per-capita usage (municipal), economic growth and energy prices (industrial), agricultural growth, farming practices, crop types (irrigation): there are plenty of variables to consider when making water demand predictions.
Because there is no way to guarantee an exact measurement of these factors in the future, there is a need to asses and quantify the uncertainty in these projections. A similar issue exists on the supply side of water planning.
Climate change, permitting and reservoir sedimentation are all areas of uncertainty that all have the potential to affect the future water supply. In the area of permitting, there is no guarantee that a specific water supplier will get the water right they were planning on obtaining.
The level of reservoir sedimentation, especially as drought lowers the level of surface water sources, can greatly affect the amount of water available for supply. Another concern and source of uncertainty is climate change.
If the rains in Texas return in earnest, groundwater and surface water sources alike will see relief. If the current drought cycle continues, or worsens, our water supplies will not replenish to levels that we expect them to and as a result we will need to go back to the drawing board with our water management plans.
For the security of future water supplies we must address the task of how to integrate all of these uncertainties into a water management plan so that the plan itself is more resilient and reliable. To start, we should consider a range of possibilities for each source of uncertainty in order to prepare for the best and worst-case scenarios. We also need to introduce mathematical techniques to quantify uncertainty.
In regards to the most important factors influencing water planning we should consider high, medium and low predictions. Where there is sufficient data, the uncertainty itself can be described using mathematical functions such as Probability Distribution Functions.
These functions should be incorporated into the planning process, allowing “likelihoods” to be ascribed to each prediction—similar to weather forecasts. Supply and demand projections should be built on an assumed range of possibilities. It should also be determined just how conservative a water plan should be: how much buffer is needed in the safety margin? All of these things can help take uncertainty into account in water planning, but the plans need to be flexible too.
There are several ways to incorporate more flexibility into water planning in order to adapt to changed conditions. Drought contingency plans should be diligently developed. Adaptive management that reacts to shifting conditions instead of remaining rigid will help address problems as they are identified. Revisiting those 50-year plans on a regular basis is important.
A suite of rapid responses to curb demand or increase supply should be developed to provide tools to effectively address immediate water planning issues in the short term. An alternative approach to add flexibility would be to simply broaden the gap between supply and demand, effectively building a greater safety factor.
That might be hard to justify to rate payers, but these are the types of corrections to the current water planning methods that will improve drought tolerance and secure sufficient water supplies for the future, even in the face of an uncertain future.
Dr. Barney Austin P.E. is the director of hydrologic services for INTERA.