| Popis: |
[1] The comment by Barsugli et al. [2009] has three main thrusts. The first deals with the water budget model we employed. In follow-up work, now published by Barnett and Pierce [2009], we revised the water budget model accounting for all the points they raise plus a few more. The results of this new work strongly support our original conclusion of upcoming, serious problems with the Colorado River water supply. A second main question deals with our assumption of 1 million acre-feet (MAF) per year (1 MAF = 1.2335 10 m)of overdrafting of Lake Mead and whether this is reasonable and representative of the combined MeadPowell system. We show that lakes Mead and Powell combined have been losing an average of nearly 2 MAF/a over the last 11 years, so our 1 MAF/a assumption is well supported by the data. Finally, they wonder about our suggestion that we begin to address the coming water problem now rather than waiting, concluding that the risks to the water supply on the Colorado are serious but that there is a window of opportunity to allow developing management strategies to address these issues. We agree that the risks are serious. But we argue that the U.S. Bureau of Reclamation’s current plan as outlined in the interim guidelines, which is to call a meeting of interested parties when Lake Mead reaches a 1025 ft (312.4 m) elevation, is tantamount to waiting until a water crisis is upon us to act. That would be both foolhardy and irresponsible; we believe the stakeholders in the Colorado River system deserve better. [2] In ‘‘When will Lake Mead go dry?’’ [Barnett and Pierce, 2008] (hereinafter referred to as BP2008), we used a simplified water balance model to examine the effects of anthropogenic climate change on the Colorado River system. In particular, we calculated the likelihood of active storage in Lake Mead becoming depleted (‘‘running dry’’), assuming no preventative action were taken, a caveat plainly stated but ignored in criticisms of BP2008. The motivation for doing this work was a United States Bureau of Reclamation (BOR) final environmental impact statement (EIS) published in late 2007, detailing guidelines for managing the main Colorado River reservoirs for the next 20 years and showing increasing water deliveries from the reservoirs for the next 50 years [Bureau of Reclamation, 2007]. Remarkably, this document took no account of anthropogenic climate change, despite the long time horizon of the report and numerous works suggesting that climate change would reduce runoff to the river (see eight examples in Table 1 of Barnett and Pierce [2009] (hereinafter referred to as BP2009)). BP2008 found that if nothing were done, there was a substantial chance of Lake Mead’s storage becoming exhausted, with ‘‘50% chance by’’ dates ranging from 2021 to 2048, depending on the particular assumptions made as to inflow to the system, delivery cuts, and the effect of climate change [Barnett and Pierce, 2008, Table 1]. There was no hint nor suggestion in the 2007 EIS that the water supply from the Colorado River faced any such risks. [3] In their comment, Barsugli et al. [2009] (hereinafter referred to as BNRPH) criticize a number of aspects of BP2008, which we address below. However, we note that Barsugli et al. [2009, paragraph ?] conclude that the ‘‘risks to the water supply on the Colorado are serious’’ and that they hope their comment ‘‘serves to put the [BP2008] analysis on a stronger footing.’’ We would be doing a disservice to the people of the southwestern United States if we did not begin by clearly pointing out our mutual agreement on this point, which stands in stark contrast to the 2007 EIS, which indicated no problems with the water supply. [4] After the reference by BP2008 was published, we received a substantial amount of feedback on that work, including somewell-reasoned criticisms of our methodology. To address these criticisms, we modified the water budget model to include more effects, including explicitly retaining all the inflows and losses below Lees Ferry, lake evaporation that depends on the surface area of the reservoirs, and a treatment of bank storage that is identical to that used in the BOR’s full model of the Colorado River (issues noted by BNRPH). The results from this improved version of the water budget model are described by BP2009. [5] All of the points raised by BNRPH are addressed in the work by BP2009, where we continue to find that humaninduced climate change, if it occurs as predicted, will seriously impair the reliability of the Colorado River water supply. However, we would like to take this opportunity to comment on the issues discussed by BNRPH. [6] First, Barsugli et al. [2009, paragraph 1] state that ‘‘the most serious of these [questionable assumptions by BP2008] is the neglect of over 800,000 acre-feet of intervening flows’’ between Lake Mead and Lake Powell. However, it needs to be understood that we compensated for this by omitting the 900,000 acre-feet (1 acre-foot = 1234 m) of annual losses below Hoover Dam. In a water balance model, only the net of inflows and losses matter, so the exclusion of compensating inflows and losses is irrelevant. This criticism is correct in the sense that the various intervening inflows and losses between Lees Ferry and the Imperial Dam do not exactly balance (they are a net positive Division of Climate, Atmospheric Sciences, and Physical Oceanography, Scripps Institution of Oceanography, La Jolla, California, USA. |
Plný text