Navajo Reservoir Operations Vol. II

Home Browse Search

appropriate station set. The distributions, expressed as a percentage of the total gain or loss by reach. were 0.0% from Archuleta to Farmington, 7.0% from Farmington to Shiprock, 58.7% trom Shlprock to four Comers, and 34.3% horn four Comers to MexIcan Hat. Using these percentages, the monthly gain or loss was computed for each intermediate station for years 1929 to 1969. For 1970 to 1993 the gain or loss was found by the difference of Reclamation natural flows. . The RiverWare model of the San Juan River Basin operates on a monthly time-step, simulating the flow at every gauging station for various depletion scenarios (current, depletion base, and various potential future projects). The model determines daily flows for the simulated Navajo Dam releases and the proposed ALP Project Ridges Basin pumping plant only. Monthly flows provided insufficient information to adequately describe the runoff hydrograph (magnitude, duration, timing, and shape) necessary in the flow recommendation process. Thus, it was necessary to temporally disaggregate monthly flows to daily flows for the San Juan River mainstem below Navajo Dam. This was achieved by a daily mass balance on the mainstem computed in a spreadsheet after each RiverWare run. The daily distribution of natural stream reach gains and losses were estimated using the difference between daily gage records. Likewise, the gaged flow records for the Animas, La Plata, and Mancos rivers at their mouths were used to disaggregate the RiverWare simulated monthly flow of each river to daily flow. Simulated monthly diversions and return flows along the mainstem were disaggregated to daily values by distributing the monthly flows into quarter month values. The distributed quarter month flows were then uniformly converted to daily flows. . Irrigation diversions, depletions, return flows, trans-basin diversions, and M&I uses were explicitly represented and modeled in RiverWare for all major San Juan tributaries (San Juan River above Navajo Dam, Piedra, Los Pinos, Animas, La Plata, and Mancos rivers and McElmo Creek). All other tributaries were aggregated into the gains and losses to the reach of the San Juan River into which they flow. The unnatural depletions from these minor tributaries were treated as direct diversions from the San Juan River. Navajo, Ridges Basin, Vallecito, and Florida reservoirs and Jackson Gulch were explicit nodes within the model and their operations were simulated according to rules. Operations of Electra Lake and all other water impoundments, including stock ponds, were ignored. However, the evaporation losses from these facilities were included as depletions from their associated streams. Several refinements were developed to compensate for peculiarities in the way the natural flow study handled some depletions and the resulting RiverWare configuration. In the natural flow study offstream depletions, remote from the mains tern and major tributaries, were treated as direct diversions from the mainstem. As a result these offstream depletions, both irrigation and non-irrigation, could call on Navajo Reservoir in the model and overdraw the reservoir during simulations. By limiting these offstream depletions to the natural gains occurring within their associated river reach, this problem was avoided. Other refinements included compensation for phreatophyte depletions along the mainstem and adjustments to lag return flows. The San Juan-Chama project was simulated following the rules of the Authorization Act. Daily bypass flow requirements in the Rio Blanco, Little Navajo, and Navajo rivers were maintained. The maximum single year diversion (270.000 aI), maximum total 10-year diversion (1,350,000 aI), and capacity of the diversion tunnels were also respected. The diverted water was stored and A-5 1'\'(T35'7'