Op-Ed: Drought Management Needs Data Science


a leaking tap on whiteBy Patrick Atwater
Special to Calbuzz

Gov. Jerry Brown just instituted mandatory rationing for the first time in the state’s history.  We’re in unprecedented territory, and this drought demands the “pioneering spirit” the governor has invoked again and again.

For starters, California needs to starting collecting and using data about its water in a manner more “pioneering” of the digital 21st century than the industrial 19th.

Consider a fundamental task with which the governor charged the State Water Resources Control Board (SWRCB): gathering basic information on water usage and actions, in order to help improve conservation efforts by local water utilities throughout the state.

The board has been collecting urban and agricultural water usage data since late last year; incredibly, most of the conservation actions are recorded in the form of machine illegible text comments, rather than as a detailed categorization of public outreach, water rate changes, and rebates like the $100+ million the Metropolitan Water District is spending to tear out laws in southern California.

Data Matters That matters, because there’s a wealth of information already contained in those comments – including 15 districts using apps — that doesn’t make it into the official summary statistics and graphs (You can see my (ugly) text parse counting the utilities that mention an “app” and my deeper dive into the SWRCB data in all its gory detail here).

There’s also basic information from other departments that doesn’t get synthesized with the ongoing usage data collection.  The Department of Water Resources (DWR) oversaw urban utilities setting water usage reduction targets for 2015 and 2020 off of long term historical water usage – far more robust than the year-over-year SWRCB usage comparisons.

So finding this sort of thing an enjoyable afternoon adventure, my friend Varun and I mashed up those two data sources to create an interactive dashboard of a geographic map side-by-side with a time series of how much utilities are over or under their 2015 targets.

The broader issue: Last year, the Delta Stewardship Council convened a wide range of California’s water data best and brightest and articulated a key barrier: “Spreadsheets are circulated with calculations performed manually, producing unnecessary cost, opaque processes, and additional risk of error.” Translation: using Excel for everything. (It’s a useful tool but not a database folks!)

droughtThe drought demands more from us. Consider some basic questions that will be critical in managing California’s water supplies.

– What’s the average price of water charged in California?  How has that changed in the last month?  In this day and age, why can’t that question be a simple database query rather than a five+ figure survey billed out to some overpriced consulting firm that doesn’t even gather a comprehensive dataset?

– What conservation actions are working and where?  The SWRCB data on mandatory restrictions is a start, although we really need to standardize those machine illegible comments on public education and rebates.

– How does the effectiveness of those actions vary across income, education, geography and other key characteristics?  What’s the best way to reach different demographics?  Maybe some groups respond better to a mailer showing how much water their neighbor uses and others like to be engaged digitally.

No Need to Reinvent the Wheel None of this is new. It’s what tech companies like Facebook do to spam you with ads or Obama’s famous data driven campaigns did to personalize outreach.  Why shouldn’t we aspire to the same level of technical excellence in dealing with the drought?

That will require integrating data collected by California’s many, many government agencies that deal with the drought and getting beyond the tribal turf battles that too often stymie progress.  Hopefully the governor’s leadership and the realities we face will be enough to shake stodgy bureaucracies out of complacency.

There’s a distinct possibility this drought won’t just go on for another year but could be the start of a new normal.  Or more accurately the reversion to the historical mean (the 20th century was abnormally wet, even before you start talking about what climate change means for the future).  Faced with that distinct possibility, we need all the water efficiency we can get.


Again we’re in uncharted territory. Data science might sound like something out of propeller-head geekdom, and asking public utilities with turf conflicts to share data might seem idealistic.  Yet ultimately that’s what just what needs to be done.  So let’s find a way to do it.

patrickatwaterSince his comeback as governor in 2010, Jerry Brown has talked repeatedly about how our public challenges demand California’s famous pioneering spirit and “show us how we depend on one another and how we have to work together.” That’s never been more true than today with the drought.

Let’s have the courage to administer the governor’s directive to the level of excellence California deserves.

Political writer Patrick Atwater is an author, entrepreneur and frequent Calbuzz commentator. He is a graduate student the Center for Urban Science and Progress.

subscribe to comments RSS

There are 3 comments for this post

  1. avatar Noozeyeguy says:

    Maybe the passengers on the critically-important multibillion-dollar “bullet train” can each be convinced to bring some water with them on their journey south?

    • avatar nslander says:

      Maybe those demanding transportation choices be limited to critically-important multibillion-dollar “freeways” and airports will finally realize expansion of those infrastructures would only accelerate climatic conditions that are, minimally, exacerbating our current drought, and stop pretending cannibalizing HSR is the only possible source of funding for any conceivable issue. Sticking with transportation, they could also re-consider the wisdom of putting one quarter of our water via almonds and alfalfa on slow boat to China, but that’s never going to happen, either.

  2. avatar JohnF says:

    Some of the Data should be easy to collect, most water bills have the number of units (gallons) used. The records should be collectable back 25 or more years. Just have the Local water companies send the records to a large server then have some Data geeks sort through it and correlate it twenty different ways. It is the 80% of the water that goes to Agriculture that I would worry about. That is not metered or measured very well. In this case the data will be either hard to get or not accurate enough to be useful. Bottom line I agree we need all of the data (information) we can get to begin to wrap our minds around this problem.

Please, feel free to post your own comment

You must be logged in to post a comment.