River basin models

From Water Wiki

Background

Models are critical to any system for managing water resources; they give us a way to predict how water supply and water demand will interact in the future. North Carolina's Division of Water Resources is pursuing a strategy to create hydrologic models for each of the state's river basins . The first few models, such as for the Catawba basin and the Cape Fear, give some indication of the pluses and minus for particular choices of models. 

Concerns

• Stationarity: the use of past river flows as a model of the future, with the assumption that there is no overall upward or downward trend in flows. Climate change makes this questionable, as discussed by these authors in Science magazine. The Cape Fear River Basin Hydrologic Model (CFHM) is based on daily average stream flows over the period of 1930-2004. Future river flows may differ from the historic record.
  
• Evaporation is a major source of water loss in all river systems, but particularly in those with large reservoirs. Evaporation increases as temperatures increase, and temperatures are increasing worldwide. The CFHM accounts for evaporation and precipitation at reservoirs, but does so based on the historic record. Future evaporation and precipitation may differ because of global climate change. A model run using otherwise conservative assumptions should also examine the effects of a 4 degrees Fahrenheit rise in average temperature by 2050. Here are some suggestions on modeling the changed evaporation. This kind of approach has been the subject of recent Congressional testimony regarding the Colorado River's flows.
  
• Global climate change: Unfortunately, the hydrologic impact of global climate change on North Carolina is uncertain. According to Dr. Ryan Boyles of the NC State Climate Office, models of global climate change are not easily applied to small regions, such as the Southeast. While Dr. Boyles has not been able to say whether global climate change would result in a particular hydrologic trend in North Carolina (e.g., dryer or wetter), Dr. Boyles has said that global climate change would likely result in increased hydrologic variability in North Carolina. Dr. David Moreau, in directing the Water Resources Research Institute in Raleigh, has summarized the climate change models as of late 2007 in terms of their hydrologic predictions, and confirmed that their predictions for precipitation changes are unclear. See further discussion at the wiki page on climate change.
  
• Other factors raise questions, perhaps significant questions, about stationarity in river modeling:
  • Increased withdrawals from surficial aquifers for irrigation supply or community water supply could mean reduced base flow levels in the future.
  • Increased impervious cover that increases the flashiness of runoff and reduces ground water recharge could, depending on its timing and the ability to store the peaks, mean reduced flows during dry conditions in the future.
  • land use changes around water in general change the water budget; for example, a transition from cleared, farmed land to forestland around streams might reduce evaporative loss and increase flows. Contra an increase in urbanized area or cleared area from forested cover.
  • While the overall effects of global warming on precipitation in North Carolina and the southeast are unclear, there is greater consensus that future precipitation patterns will change towards less frequent, more intense events. This could itself have profound impacts on water availability. "The distribution of lengths and timing of dry spells are of as much concern as actual precipitation and evaporation values in controlling reservoir levels."^[1]
    
  • Long term sedimentation. Here is an article by authors at Duke's Nicholas School exploring ways to account for long-term sedimentation changes in hydrologic modeling.
• The CFHM models withdrawals and discharges based on seasonal patterns for each system included in the model, and applies a unique and constant factor to each withdrawal and discharge based on an assumed annual average. This method does not take into account the reductions in withdrawal and discharge that might occur as a result of water shortage restrictions imposed during drought conditions.
• For western and piedmont North Carolina, the drought of 2007-08 saw some flows that were even lower than in 2002. The models should incorporate these '07'-08 flow years.
• we should have some way of modeling what will happen if there is much increase in use of reclaimed water
• does the model work to assess what would happen with the skimming of high flows? if not, I think that is an important limitation that should be addressed in future river basin models

Other models of interest

Here are some other hydrologic models with interesting features, such as being aimed at educating local water decision makers.

Notes

1. ↑ PETER J. ROBINSON, INTERNATIONAL JOURNAL OF CLIMATOLOGY, VOL. 17, 983±996 (1997) CLIMATE CHANGE AND HYDROPOWER GENERATION