The CaDC provides an integrated suite of open analytics to support water managers in meeting their reliability objectives.  These tools fall into a couple key categories, which are summarized below and then examined in greater detail:

  1. Navigating statewide efficiency policy -- by having data integrated across utilities, CaDC analytics can be used to inform and understand the implications of prospective statewide policy using actual data that reflects the diversity of local circumstances and conditions.  

  2. Evaluating program effectiveness -- conservation and water efficiency programs have historically been a “feel good” action. By providing the foundation to rigorously evaluate demand management programs, data supports water managers in achieving reliability through an approach that balances new supply development and demand management.  

  3. Supporting service reliability -- increased water scarcity translates into lower water sales and often decreased revenue.  The CaDC has developed a new rate modeling tool to illustrate at a high level the impact of prospective rate shifts on customer bills and district revenue.  

Fork me on GitHub
“The new Efficiency Explorer tool is exactly what we needed to help us work more efficiently. We are now able to work quickly and easily with our water efficiency data to quickly obtain information, giving us more time to focus on helping our customers reduce their water use. This is a perfect solution for water agencies.”
— Fiona Sanchez, Irvine Ranche Water District

The CaDC Efficiency Explorers provide interactive dashboards that water managers can use to understand the implications of the Governor's new residential efficiency standards.  The video below demonstrates that functionality at a utility and neighborhood level.

Landscape area is a key missing data point in calculating these efficiency standards statewide.  The CaDC has partnered with Andrew Marx (CGU), who has developed a new algorithm to classify irrigable area using the National Agricultural Imagery Project (NAIP).  This imagery is free, publicly available and in 2016 is available at a higher resolution (60 cm versus 1 meter in 2014). 

NAIP imagery and resulting classifications.  Courtesy of Andrew Marx (CGU).  

NAIP imagery and resulting classifications.  Courtesy of Andrew Marx (CGU).  

This algorithm could be redeployed to other, higher resolution proprietary data sets.  In addition, the CaDC has been working closely with leading vendors on how to develop those landscape area and other measurements as part of an integrated, phased approach to statewide water efficiency.  That data offers key context, not only for setting efficiency standards, but for providing tools to support water managers to measure program effectiveness and understand what works in their unique service area in order to actually meet those efficiency targets.  

This sort of interactive visualization can also be used to benchmark CII water use efficiency through more granular customer categories like restaurants or warehouses rather than commercial customers broadly.  See here for an example of this approach deployed for the City of New York.  Several CaDC staff helped develop that tool.

The Data Collaborative has applied statistical techniques to estimate the actual water savings achieved through turf removal rebates. The results of that approach (using bayesian structural time series) are shown below and are described in full in the CaDC study published by KDD, which is available here.*

Predicted monthly savings for each household in the data set. The dark green line corresponds to median savings. Seasonal variation leads to swings in average savings from -1.5 to -2.7 gallons per square foot.

Predicted monthly savings for each household in the data set. The dark green line corresponds to median savings. Seasonal variation leads to swings in average savings from -1.5 to -2.7 gallons per square foot.

For the CaDC study of the water savings associated with turf removal, mean predicted savings for single-family residential accounts are estimated at 24.6 gallons per square foot per year for the households used in the study. The study utilized a data set of 545 unique single-family residential turf rebates across 3 California water utilities, totaling 635,713 square feet of converted turf grass.

The blue line shows the observed usage for one household. The red line shows the expected usage based on the consumption of households with historically similar usage patterns. The vertical dotted line represents the time the turf removal was performed. A 29 percent reduction from expected water use is visible after the removal.

The blue line shows the observed usage for one household. The red line shows the expected usage based on the consumption of households with historically similar usage patterns. The vertical dotted line represents the time the turf removal was performed. A 29 percent reduction from expected water use is visible after the removal.

When this technique is combined with automated monthly data ingestion, it provides the ability to generate ongoing estimates of program savings to adaptively manage the historic investment in turf rebates. This sort of approach also opens the door to more advanced analyses such as outlier detection of accounts with suboptimal water savings.  

More broadly, the approach can be applied to estimate the causal impact of any targeted conservation program by comparing the usage patterns of affected and unaffected customers and how these patterns change after a program takes effect.  Several other studies are ongoing and we are working with the CaDC technical working group to develop an expanded version of this analysis.

Of course, the ultimate effectiveness of this historic investment in turf rebates will come down to two key things: 1) how long the lawns will stay converted and 2) how many neighbors were influenced to change their landscape.  The chart below shows the cost effectiveness of the turf rebate program broken out by those two axes.  The calculations used to generate this chart are available here.

The ultimate effectiveness of the turf rebate program will come down to those two factors: lifespan and peer effect.   If landscapes do not stay converted and lawns stay the default in ornamental landscaping, the turf rebate program will go down in history as not cost effective.  

Yet if this historic investment is the tipping point in making California friendly plants mainstream, it will go down in history as a visionary and extremely cost effective program.   So below is a strategy on continuing the turf market transformation momentum to deliver on exactly that.  

Turf rebates have succeeded in their goal: converting early adopters away from water thirsty lawns to California friendly landscapes.

Yet even with the historic investment in turf rebates, we are only at approximately 1.6%  of turf market transformation in terms of the total square footage of turf.  Smart conservation pricing and public education are cost effective ways to continue the momentum in making California friendly plants mainstream. 

The goal of course is not to eliminate lawns.  Yet ornamental landscaping like street medians does not need to be a water thirsty lawn.  In fact, per the latest Governor’s executive order lawns in street medians cannot be irrigated with potable water.

*CORRECTION: An early version of the translation of the water savings associated with turf removal into a dollar per acre foot value had a unit conversion error and yielded an incorrectly low dollars per acre foot figure.  This has been corrected in the graph shown and final paper linked above.  Assuming a 30 year lifespan, no peer effect and a 5% hyperbolic discount rate, the correct translation of 24.6 gallons per square foot into $1422 dollars per acre foot of water saved from a $2 per square foot turf rebate. 

With declining water sales, water managers have lost over $675 million of revenue in the current drought.  The CaDC rate comparison tool provides a quick way to see the implications that a rate shift or a drought surcharge would have on revenue and typical customer bills. The tool quickly illustrates the impact of a shift before a utility hires a rate consultant and goes through a full, labor intensive Prop 218 process.  

Future work: This tool currently supports single family residential customers and is being expanded to the entire spectrum of utility customer classes.  In addition, future iterations of this tool will support scenario analysis on changes in water sales and ultimately the vision is to incorporate state-of-the-art econometric models estimating the impact of price on water demand from our academic partners.  Improving that estimate of the price elasticity of water demand is a key priority of the CaDC research agenda.

Predicting the future

Providing a short term demand forecasting tool to balance peak loads

Faced with a historic drought and future uncertainty, California water managers need all the tools they can get to ensure water reliability. Short term demand forecasting has a long track record of helping manage peak loads in energy, and this new tool created by DataKind brings that approach to help manage recycled water demand peak loads and reduce the need for potable makeup water for irrigation watering. Click the picture below of the dashboard to see the latest version of the tool.

Please click on the picture above to see the depicted dashboard predicting recycled water demand in Moulton Niguel Water District. 

Please click on the picture above to see the depicted dashboard predicting recycled water demand in Moulton Niguel Water District. 

That tackles a key need for Moulton Niguel Water District in South Orange County, which aside from recycled water, is 100% reliant on expensive imported water supplies and serves as a leading example of how improved use of data can ensure water reliability.   The graphic below summarizes the value of this approach for Moulton Niguel Water District and the public benefit it provides to California.  The underlying code is open source and available here.

Produced in partnership with the following incredible organizations: