Urban Water Use Objective Allocation Optimization with Beaumont Cherry Valley Municipal Water District

Eagle Aerial Solutions (Eagle) worked with Beaumont Cherry Valley Municipal Water District to update their 2018  parcel measurements. Updating the irrigated landscape area measurements of parcels added since 2018 will improve the accuracy of the data the State Water Resources Control Board (SWRCB) is using to calculate the water use allocations.  

Having worked on the original 2018 Landscape Area Measurement data, Eagle has the expertise to update data for new construction or misclassified parcels.  Eagle has developed a scalable methodology to update all of Beaumont Cherry Valley’s parcels with 95% accuracy. This methodology is equal to or superior to the original Landscape Area Measurements and is approved by the State Water Resources Control Board.

Results:

1,884 new residential parcels with an estimated Irrigable Irrigated area of 3,743,075.62 square feet

1,543 misclassified parcels with an estimated Irrigable Irrigated area of 5,149,970,.57

Total of 3,427 parcels that were not included in the original LAM report with an estimated additional square footage of 8,893,046.19

Landscape Area Measurement Data:

• Original:  42,762,231 square feet (Irrigable Irrigated area Cell B30 in template)
• New:  8,893,046.19 square feet
• New Total:  51,655,277.19 (21% increase in Irrigable Irrigated area)
• Original Allocation:  7,829.48 acre-feet
• Newly Constructed Landscapes get a Landscape Efficiency Factor of .55
• New Allocation:    8,385.82 acre-feet (additional 556.34 acre-feet, 7% increase)    

Next Steps:

Eagle Aerial Solutions will update the parcel maps for BCVWD and add a new GIS layer into their WaterView account.  The new layer will be provided to BCVWD and included in an official package to be submitted to the State Water Resources Control Board for approval and then added in to the calculation of the new consumption targets.  This will give BCVWD an updated and more accurate picture of its current water use and assist in future program planning.

Background

Reporting

California water-use-efficiency legislation requires water agencies to meet annual residential water use targets.  These targets are calculated using parcel and customer demographic data such as the amount of landscape irrigated in square feet, weather data, and number of people per household.  The data the State Water Resources Control Board uses to calculate the targets is based on aerial imagery from 2018.  Since BCVWD has had significant growth since 2018, the District has engaged Eagle Aerial Solutions to update the parcel data.  This new data will improve the accuracy of the SWRCB target calculations and give BCVWD a more accurate picture of its residential water use.

BCVWD is required to report consumption for both Residential and Commercial customers on a fiscal year basis.  This requires tracking and compiling consumption for Residential customers as well as those commercial customers who have Dedicated Irrigation Meters.

Eagle Aerial Solutions

Eagle Aerial Solutions has over 30 years of experience working with aerial imagery and has extensive knowledge of the requirements of the legislation. Eagle Aerial Solutions and its partner NV5 were the contractors who provided the original data to the state and have the expertise to update the data to the state’s specifications.  Eagle staff also have extensive expertise in working with residential and commercial parcel data.

In order to assist agencies with water-use- efficiency and legislative compliance, Eagle Aerial Solutions developed WaterView, a simplified GIS platform that will allow staff to view customer consumption on a monthly basis and to easily see which customers are using more water than their estimated budgets.  WaterView also provides analytical tools to help identify trends in consumption, the ability to track select groups of customers, and the ability to export lists of customers for focused outreach.  Finally, WaterView is continuously updated with the latest requirements from the legislation so staff do not need to track it themselves.

• Eagle Aerial Solutions (Eagle) supports clients in meeting compliance targets related to California water-use-efficiency legislation (AB1666 and SB606).
• Outdoor Residential Targets are based on the square-footage of Irrigable Irrigated area measured in 2018.
• This seven year-old data does not include any New Residential Construction since 2018 and might have missing parcels and/or misclassified parcels that have Irrigable Irrigated area that should be included in the overall allocation.

Other Square footage not included in original LAM data

• Residential Special Landscape Area
• Residential Parcel Extensions (when provided by DWR) ‍

Exhibit 1: Methodology

State Water Board guidelines allow for updates of the 2018 baseline landscape area measurement (LAM) data.  In accordance with these guidelines, Eagle Aerial Solutions developed the following methodology to update BCVWD’s parcel maps.  

Methodology for Estimating LAM Values for New Residential Parcels

Eagle Aerial Solutions will update the following parcel data for BCVWD:

• New parcels not in the 2018 DWR Parcel dataset that meet the DWR residential land use classifications OR have a residential meter on them
  • Inside the agency boundary
  • Outside the agency boundary
• Misclassified parcels existing in the 2018 DWR dataset that DID NOT have a DWR Residential land use classification but now do have a DWR qualified Residential land use classification
• Residential parcels existing in the 2018 DWR dataset with a Residential classification that have very low to no Irrigated Area but have a Residential meter on them

To assign Landscape Area Measurement (LAM) values to newly developed residential (RES) parcels that do not appear in the 2018 baseline dataset, we employ a quantile-based proxy estimation framework augmented with a targeted refinement for large-lot parcels. This allows us to produce statistically consistent, scalable LAM estimates without requiring manual measurement or costly remote sensing.

1. Create Parcel-Size Cohorts (Quantile + Upper-Tail Refinement)
‍

We begin by generating parcel-size quantiles for the majority of existing RES parcels, creating equal-frequency stratification bins that group parcels with similar lot sizes.

Because parcel size is highly skewed—particularly at the upper end of the distribution—we then apply a targeted refinement for the upper tail. Specifically, after forming the initial quantile bins, we manually subdivide the largest parcel-size segment(s) to create additional cohorts.

This refinement provides:

• Better resolution among unusually large or outlier parcels
• More accurate LAM estimation for parcel classes whose landscape characteristics diverge from the general RES population
• Improved model fidelity for districts with heterogeneous residential development patterns

The resulting set of parcel-size cohorts balances statistical rigor with practical modeling sensitivity.

‍

2. Calculate Representative LAM Values per Cohort
‍

For each parcel-size cohort (including the refined upper-tail segments), we compute the average LAM component values:

• II — Irrigable Irrigated Area
• INI — Irrigable Not-Irrigated Area
• NI — Not-Irrigable Area

These cohort-level averages form a set of reference LAM profiles that characterize landscape allocation patterns for parcels of comparable size.

3. Map New Parcels to the Appropriate Cohort

Each new RES parcel is assigned to the cohort whose parcel-size range it falls within. The manually refined upper-tail cohorts ensure appropriate classification for large-lot parcels that would otherwise be overgeneralized in a pure quantile model.

4. Inherit Cohort-Level Baseline LAM Values

Once matched, each new parcel inherits the representative LAM profile (II, INI, NI values) for its assigned cohort. This approach yields:

• Consistent and reproducible estimates
• Statistical alignment with real-world parcel patterns
• A scalable method suitable for annual updates

This process provides a defensible, data-driven approximation of baseline LAM values in the absence of parcel-specific measurements.