Missing water

There are about 334,000 irrigated acres within 2.5 miles of the streams and tributaries of the Republican River in Nebraska.  On average, they pump about 295,000 acre feet of water each year from the aquifer.  According to the DNR, if all of those wells stop pumping water out of the ground then, by about year six after the wells are turned off, there will be an additional 50,000 acre feet of water in the stream.  The 50,000 acre feet number comes from a Model run done by the DNR in November 2006 and reported to the NRDs in January of 2007.  The worksheet is titled 0746MdDrt_ImpactReductionsSummary.xls   Sub-Worksheet 3-NRD. Column X was used, as that is the column that shows the benefit to the stream if all Quick Response wells are turned off.  You can see it here.

The question is, where is all of the missing water accounted for?   According to public statements made by Ann Bleed (DNR), somewhere around 90% of the water pumped out of the aquifer in the Quick Response area would be in the stream, if it had not been pumped.  Yet the Model shows that only about 17% of the water pumped would have been in the stream.  This is a huge discrepancy.   If Ann Bleed is correct, then the Model should show the blue line approaching the red line after a number of years.  But, it doesn't.  So, is the Model wrong?  Or, is Ann Bleed wrong?  Or, are both wrong?

I provided this information to several NRD managers, to every Senator on the Natural Resources Committee, and to the DNR via the NRC legal counsel.  I have not received any direct response; however, there have been a couple of comments about the concept.  The explanation from the DNR goes like this.  According to the DNR, the reason the water does not end up in the stream when the wells are shut off is because evapotranspiration goes up and consumes the water.  So according to the DNR, the water disappears from the system either because corn uses it or if there isn't any corn, then trees and grass use it.  So it doesn't matter if we turn the wells off or not, the water is evaporated away either way.    That would mean we are spending millions of dollars to reduce irrigation and according to the DNR, most of that saved water will be consumed by grass and trees and will not be available for compliance with Kansas.

In realty, what has happened is that the people who made the Model put a cap on the benefit that can be derived from shutting off irrigation wells.  The cap is at 17% of what was pumped.  The average well may pump 130 acre feet of water each year.  The DNR has said that about 90% of that water would have been in the stream if it had not been pumped but the Model says that after about the fifth year of the well being off that there will 22 more acre feet of water in the stream each year forever.  That doesn't match reality but it is what the Model says will happen.

The effect of this is that either the wells don't affect the stream as much as the DNR says they do or that the Model says more wells have to be shut than really have to be shut off in order to cause compliance.   Either way, there is a serious problem caused by the 17% cap on the benefit of wells being shut off.

Below, I show how the 295,000 acre feet was determined.  It doesn't really matter what this number is.   Move it up or down by a large amount.  There is still a lot of missing water.

Pumping by Quick Response (QR) acres is estimated by taking total pumping and prorating it across the applicable acres.  It is important to note that pumping varies by a large amount each year, and this is an average.

Another way of thinking about it:  If all of the water pumped from the aquifer would eventually be in the stream, as Ann Bleed has repeatedly said in public,  then couldn't we do a proof, as in algebra, and add up all of the depletions and see how much water the Model says would be in the stream?  In other words, if the depletion caused to the stream by a cell is 80% over 50 years and the well pumped 100 acre feet a year over those 50 years, then it would have pumped a total of 5,000 acre feet and the Model says that depletions are 80%, or 4,000 acre feet.  In year 50, the depletion to the stream by the accumulated 50 years of pumping might be 92 acre feet.  In year one, it might be one foot; in year two, 1.2 foot; etc.  Each year, the depletion caused by prior pumping would increase.  By year fifty, the depletions might be 92 feet in that year. 

Now add those computed depletions for every cell in the Basin for the year you want to look at and see if the total is less than the amount of water that was ever historically in the stream.  If the total is more, then there is a problem with the Model and it needs to be recalibrated.  If it is less, then it is possible the Model is correctly making its estimates.

To my knowledge, this check or proof of the system has never been done.  I suspect -- but don't know -- that such a check would find a problem, as the depletions suggested by Ann Bleed would be impossible.  The depletions suggested by the Model would be possible, but those depletion percentages are very different than what Ann Bleed suggests and also have the problem shown in the chart above.