I had the opportunity to chat with Felicity last week as he prepared the story. Of course there is so much research that goes into putting these together and only so much that goes into print (that's why we all have blogs...). Here, in brief, are some of my thoughts on the topic, many of which I shared with Felicity:
Agricultural water use is not equal to the sum of groundwater pumping and surface water delivery. Crop consumptive use (the actual water evaporated into the atmosphere from crops, trees, vines, etc.) is less - sometimes much less - than the water applied. What happens with the rest? It runs off into a river or - in most of California's agricultural regions - it recharges a groundwater basin. Putting a meter on a well does not "measure" the farms groundwater "use" as 10% to 50% of that may just percolate right back to the groundwater.
Then how should farm water use be measured instead? A few years back, CALFED convened an independent panel on Appropriate Measurement of Agricultural Water Use, which I had the honor to be part of. Our final report highlighted key options to go about measuring both surface water use and groundwater use in agricultural settings:
---Independent Panel Recommendation----
* Employ more precise methods to compute and report net groundwater usage to the State. [...] Accordingly, the Panel recommends that the State employ more precise methods—specifically, continuous regional characterization of groundwater volume to compute net usage using two methods simultaneously:
(1) development of detailed sub-basin hydrologic balances; and
(2) the water table/specific yield method (which requires measuring water levels rather than groundwater extraction).
This approach represents a substantial change from current practices. Footnote: Felicity's article mentions the Pavley bill (SB 122), which would be a big first step towards doing just that: monitor water levels throughout the state (It's a reincarnation of previous year's bills, all of which - including this one - have been championed by the Groundwater Resources Association).
* When water transfers involve groundwater substitution, the groundwater wells directly involved in the transfer require some form of continuous measurement,
monitoring and frequent reporting.
* This definition [of groundwater measurement] should in no way be considered to preclude or limit higher standards of groundwater measurement that may be deemed necessary by entities with legal jurisdiction over groundwater management, including local agencies or authorities, to meet site- or condition-specific needs.
* Crop Water Consumption Measurement: Measure using satellite-generated remote-sensing. Current approaches to measuring crop water consumption rely on indirect methods applied infrequently, a practice that means state estimates of crop consumption—a significant portion of California’s total water use—are not validated
and could include significant error. The Panel’s recommended approach—using satellite-generated remote sensing to measure crop consumption—is expected to
yield significantly better estimates than current practices. [...] and would have no direct impact on water users.
-------------------------------------------------
On this latter approach: my former grad student Alec Naugle and former postdoc Nels Ruud, a few years back, used a straight-forward water balance approach, based on monthly precipitation, monthly surface water deliveries to individual water/irrigation districts, monthly field-by-field crop consumptive use, and soil moisture tracking, to estimate monthly groundwater pumping for each of about 10,000 individual fields in southern Tulare County from 1970-2000. This approach may be as good or better as any well extraction metering approach to get a handle on groundwater usage. Detailed maps are available in our 2003 final report and the methodology and regional results with examples are published in a Journal of Hydroloy article (Ruud, N.C. et al., 2004) (see the bottom of our project page for these and related pubs). Our estimated field-by-field groundwater usage, aggregated to the district and regional scale, compared well with the water level fluctuations observed in that basin - confirming the importance of water level measurements as one of the cornerstones to understanding net groundwater usage.
David Maidment, Univ of Texas, told me on his visit to UC Davis today that around the same time our Water Panel convened, he was chairing a committee on the National Academy of Sciences that published "Estimating Water Use in the U.S.: A New Paradigm [..]". Chapter 5 of this report describes a "stratified random sampling" approach, a technical term for the kind of smart polling done in presidential elections, to estimate actual groundwater extraction. The approach was tested on the Arkansas database, containing 36,000 groundwater withdrawal points. The conclusion of the exercise was that "the single largest withdrawal points should be measured, and a random sample of 109 points (approximately 0.3 percent) should be selected from the remaining 36,052 groundwater withdrawal points. Hence, with groundwater withdrawals for irrigation, which account for over 80 percent of all withdrawal points in the state database, random sampling is sufficient for statewide water use estimation." (p. 94). In other words, a small - smartly selected - subsample of existing wells may be sufficient to get a good estimate of actual extraction. There are some caveats to doing this in the absence of the kind of database available in Arkansas, but far from everything needs to be measured to estimate groundwater withdrawals at the scale necessary to manage groundwater.
Overdraft, of course, has been one of the key topics that Felicity and I spoke about. Given the drought, which has led to recent drops in water levels, overdraft is popularly looked at as the imminent result of the state's lack of a rigorous groundwater permit system. Looking at well level records of past droughts, I have found that water level drops of 10 feet to more than 30 feet per drought year are not uncommon in the San Joaquin and Tulare Lake Basin. The repeated cycles of drought, during which groundwater is "overdrafted" and wet periods, during which groundwater levels recover can be nicely seen in the above-mentioned J. of Hydrology article. The below shows the regional groundwater storage starting from 1970.
[Note: this is the net change - in cm height of water - in water storage across the entire project area - about half of southern Tulare County on the valley floor. 30 cm is one acre-foot per acre. In terms of water levels, 3 cm net water storage change correspond approximately to one foot in water level change.]
Depending on which year you start and end your analysis, you may see "recovery" (e.g., 1977-1998), "no change" (1970-2000), or "overdraft" (1970-1995). Given that the large groundwater storage swings occur over the course of a decade; and given that it's only been 4 decades since we have the state and federal water projects in place, we often have simply not a long enough time series to establish that a basin is indeed in a period of "long-term net decline in water levels", which is a broad definition of overdraft. More importantly, a three- or five-year decline by itself does not establish "overdraft", it may just be part of good (intentional or not) groundwater management. This is not to say that there are basins with clear overdraft, only to demonstrate the difficulties and shortcomings in determining overdraft with relatively short time-series of water level measurements.
Groundwater regulation, fueled substantially by the last three years of drought, has been in the California news quite a bit ever since Cathryn Freeman from the CA Legislative Analyst's Office (LAO) suggested it in two reports last year, a water primer and a brief on groundwater management, and recently reiterated in "Water Rights: Issues and Perspectives": that the state "regulate" groundwater:
"However, successful implementation of this solution is hampered because groundwater use is generally not regulated or monitored at the state level (in contrast to surface water). In addition, local groundwater management does not take into
account statewide water needs. Finally, groundwater quality is not protected under state regulation as comprehensively as surface water quality. When contaminated, groundwater loses its potential to serve as a water supply source. Recommend Statewide Groundwater Rights and Quality Permitting System. For the reasons stated above, we recommend that the Legislature establish a state-administered water rights system for groundwater." (p.69 of the LAO water primer).
I have some open questions on the use of an outright permit system for groundwater:
* as California may be the last state without a groundwater permit system, is it really doing any worse with respect to groundwater management than the other 49 states?
* all of the States overlying the Ogallala / High Plains Aquifer regulate groundwater - has that stopped overdraft from that aquifer system?
* by which measure do we judge "good" and "bad" groundwater management? The popular answer would probably be: whether or not there is overdraft. But when exactly is a basin in overdraft - see above image and discussion?
* how much groundwater would be allocated to permits? There are examples from adjudicated basins - typically based on "safe yield" estimates. How would that be determined under conditions of climate change?
* Catherine, in her LAO report, laments the shortcomings of the surface water rights system, which IS based on permits. Why should we apply the same system to groundwater? Management is still left to local agencies. My hunch is that the existing, locally controlled system provides much more flexibility as is, as long as the "bottom of the barrel" (lowest allowable water level) is defined clearly and consequences of exceeding that depth are set out clearly. The idea of "permits" seems simple enough to grasp as a solution, but is it really effective in managing groundwater? In and by itself, it doesn't do anything to manage groundwater, especially since groundwater - unlike surface water - must really considered to be a bank/storage reservoir, rather than an incidental stream of water that is either there or not there at any given time, to be distributed and used at that time.
Well, well, we will certainly talk a lot more about this at the upcoming Biennial Groundwater Conference, 6-7 October 2009, in Sacramento!
Ah, and before I forget: Felicity's colleague Alicia inquired about comparing state groundwater usage in the U.S as part of Felicity's story. This is what I wrote her back:
The USGS Circular 1268 would be the best resource for making any
state-to-state comparisons that I know of.
On this following page, you find text that says "California accounted
for 18 percent of total groundwater withdrawals." [note that the USGS
is now changing the spelling of 'ground-water' to the much more common
spelling 'groundwater']
http://pubs.usgs.gov/circ/
A map is shown in the lower left of their Figure 2:
http://pubs.usgs.gov/circ/
The corresponding numbers are tabulated by state in their Table 1
under "Ground Water"-"Fresh" (2nd column from the left):
http://pubs.usgs.gov/circ/
While these numbers are for the Year 2000, newer numbers would not
necessarily add better or more updated information. Groundwater
pumping in states like Texas and California varies quite dramatically
depending on whether we have a wet year, a normal year, or a very dry
year. In California, groundwater pumping varies from approximately 10
million acre-feet per year in wet years to nearly 20 million acre-feet
per year in dry years. See, for example, the California Water Plan
Update 2005, Figure 1-10, lower panel, in the following pdf file:
http://www.waterplan.water.ca.
Groundwater pumping is shown in dark burgundy or purple color for a
wet (1998), normal (2000), and dry (2001) year. That means, in any
year, groundwater contributes from approximately one-third to well
over one-half of the total fresh water supply for agriculture and
urban uses in California.
No comments:
Post a Comment