Model Problems

The computer simulation that is used to measure Nebraska's compliance with the Agreement it has with Kansas regarding the Republican River is called the Model. I have spent many months reviewing the Model. There are several things that appear to be problematic.

Problem	Brief Description	Effect
Missing Water	About 250,000 AF can be removed from the system each year without affecting the stream. This conflicts with what the DNR tells everyone.	Model potentially in error by a huge amount. Potentially more wells are turned off than necessary or groundwater irrigation has much less of an effect on the stream than what the DNR reports..
Allocations, No Rebound	Even though precipitation has been nearly normal for several years, the allocation has not rebounded.	Reveals that precipitation is not the primary determinant of Model allocations but, instead, stream flows are the primary determinant. Thus, prayers for precipitation should focus on asking for rain on the stream, as other areas don't really matter.
Input Data (Pumping)	Is accurate data going into the Model? The data released on the RRCA (Republican River Compact Administration) web site in 2005 showed pumping in the Basin at a rate three times greater than what was actual.	If the inaccurate inputs are still used in the Model, then depletions to the stream caused by groundwater lag effects would be overstated. Hence, Nebraska would be out of compliance by a greater amount than it really is.
Input Data (Base Flows)	Is accurate data going into the Model? The data released on the RRCA web site in 2005 showed base flows at several gages to be greater than stream flows. That is not possible.	If the inaccurate inputs are still used in the Model, then the historical flows coming from groundwater are significantly overstated. This incorrectly results in a large drop in flows being attributed to groundwater pumping when compared to historical flows.
Inaccurate Aquifer Level Predictions	When the Model aquifer level predictions are compared to actual levels, the Model is wrong on about 56% of locations measured. It is wrong mostly in the eastern end of the Basin. The Model says the aquifer went down much more than it actually did. This causes the Model to overestimate base flow depletions caused by groundwater pumping.	Because the system makes a number of inaccurate predictions, it is unreliable.
Conservation	The effects of conservation were intentionally excluded from the system. The effects of conservation are assigned to groundwater pumping. The DNR agrees that conservation is not in the Model but denies that groundwater pumping has been assigned the effects from conservation. However, a review of the numbers shows that conservation effects ARE assigned to groundwater irrigation. Conservation was intentionally omitted from the formulas because the DNR did not want any reduction in conservation even though it is the primary cause of the problem. To make the formulas balance something else had to take the blame. Groundwater irrigation took most of it. As a result when irrigation wells are turned off they stream will not increase as the Model predicts.	Groundwater pumping is assigned much greater responsibility than what it actually causes.
Base Flow Calculation Methodology	According to the Model, over 50% of the base flows are caused by precipitation events in the alluvium within a few days of the precipitation event. This means about half of the base flow values used by the Model are not aquifer related. As a result of the way base flows are calculated groundwater pumping is assigned a much greater responsibility than it causes.	Changes the cause of stream flow depletions by assigning to groundwater pumping the effects of vegetation and conservation to reduced base flows.
Base Flows and Dams	The Model says that the dams caused a permanent 50% reduction in the amount of water oozing from the aquifer when they were constructed. The dams were put in place before groundwater irrigation became common. Is this drop in base flows attributed to groundwater irrigation? Is it accurate? It does not appear to be correct.	Unknown effect. Would need to run the simulation with accurate data in order to see what the effect really would be.
One Layer	The Model is only one layer.	This prohibits the software from differentiating between the alluvial and deep aquifer. It also means the Model cannot recognize the fact that much of the Ogallala formation is disconnected from the Republican River.
Why the problems persist