How to read the water

Reservoir storage and elevation

stock / elevation · Dated daily instant; month-end for monthly mode

Native acre-feet and feet; no interpolation. Use DATE TIME, not OBS DATE.

Lake Mead: 6,816,799.80 acre-ft, Oct 12, 2026; a stock, not annual consumption.

Source: Reclamation operations and DWR CDEC sensor 15 / 6

Provisional; eight Reclamation histories lack Oct1–12,2025. Don Pedro has three missing days.

Reservoir capacity and percent full

reference capacity; ratio of compatible stock · Source-specific stable capacity reference

Stored / stated capacity ×100; federal AOP live+vacant denominators and results marked approximate.

Mead 6,816,799.80 / 26,120,000 ×100 = 26.1% at joint-use full pool.

Source: CDEC RES, Reclamation capacity sheet and AOP Table1

Not dependable supply. Seven federal reference capacities are approximate; no precision ranking.

River gauge discharge

flow rate · Observation timestamp; not integrated annual volume

Retain native cfs, timestamp and provisional flag.

Lees Ferry 7,500 cfs at Oct13,2026 13:00 −07:00.

Source: USGS site 09380000 parameter00060

Gauge discharge is not legal apportionment or consumptive use.

Station snow water equivalent and WY peak

depth of water in snowpack · Latest dated observation; WY2026 peak is maximum valid daily station SWE Oct1,2025–Sep30,2026

Retain inches. Reject negative sensor readings; blanks stay null. Compare the observed WY2026 maximum with the source-published median of seasonal peaks from the NRCS SNOTEL Snowpack Update Report, 1991–2020. Preserve shorter normal-record counts, including Agua Canyon’s26years. Daily medians remain separately retained, never relabeled as seasonal peak medians. Peak percent of normal is observed season peak / published station median seasonal peak ×100, rounded to a whole percent; it is not a basin index.

Berthoud Summit observed WY2026 peak12.2inches on Mar17,2026 beside23.2inches published median seasonal peak, 1991–2020; normal peak date May03. Latest 0.1 inch Oct13,2026. The peak benchmark is53% of the published23.2-inch median, not percent of the15.9-inch daily median on Mar17.

Source: NRCS WTEQ value / median_1991, SNOTEL Snowpack Update Report seasonal peak medians, and CDEC daily sensor3

NRCS seasonal peak medians are published source statistics, not maxima of daily medians. No CDEC median acquired. Peaks with missing days are observed maxima, not guaranteed seasonal maxima. Stations do not represent whole basins.

WY2026 monthly exploration

exact month-end stock / SWE snapshot · Twelve months Oct2025–Sep2026

Take only the record on the calendar month’s last day. Missing day gives No reading. No nearest day, averaging, interpolation or annual/12.

March2026 at Oroville uses Mar31,2026 storage; the card and time caption identify month-end.

Source: Retained daily reservoir and snow series; precomputed per-entity summaries

Annual use and legal context retain their own dates and muted presentation. Geography stays dated. No monthly reservoir elevation or consumptive-use series is implied.

Lower Colorado entitlement and actual use

Beneficial consumptive entitlement / approved use / source-defined consumption, or explicitly labeled diversion · Current CY2026 legal/operational context beside actual CY2025; dated source releases

Transcribe exact recipient/quantity and preserve acre-ft. Consumption is mainstream diversions less qualifying returns, not gauge flow. No sum across state and user totals.

Arizona: 2,800,000 basic less 512,000 shortage+DCP = 2,288,000 acre-ft CY2026. CY2025 consumed 1,910,524. Different years: no delta. CY2025 fully adjusted availability 1,910,524 less same-year consumption 1,910,524 = 0 acre-ft.

Source: Reclamation CY2025 Accounting Report Tables4–6,11; 2026AOP and Part417 approvals; entitlement listings

State outline is orientation, not a spatial adjudication of rights. Fully adjusted availability includes conservation/ICS adjustments; the accounting balance is not a violation or physical-shortage measure. IID’s current approval includes LCWSP and is context only versus older actual. CAP balance is variable. Big Bend contract diversion is not consumption.

SWP / CVP allocation percentages

announced allocation percentage, not delivery · 2026 announcements with issue dates

Retain source denominator: requested SWP supply vs CVP south-of-Delta irrigation contract total.

SWP45% requested supply; CVP28% contract total. No cross-project difference.

Source: DWR May15,2026 and Reclamation Aug11,2026 notices

Later supersession clearance incomplete; annual announcement is not a monthly observed value.

City supply portfolios

reported annual volume categories or rounded narrative shares · LADWP FY2024/25, Phoenix CY2025, LVVWD report-on-2025 context

Keep original periods and denominator. Never normalize a range or infer missing bar values.

Phoenix surface97–99%, groundwater about2%, retained independently. LADWP groundwater6,107 and recycled13,241 acre-ft.

Source: Agency annual / water-quality reports

LVVWD90/10 is not an audited volume budget or every SNWA member. LADWP complete shares not tabulated. Recycled environmental reuse excluded.

Mexico treaty and reduction context

scheduled / allocated delivery, reduction, recoverable savings, conservation · CY2026 vs adopted CY2027–2028; each has effective dates

Keep published initial1,352,595 acre-ft unchanged. Separate 50,000 reduction, 30,000 recoverable savings, and at least67,000 additional conservation in2026.

Minute334: 250,000 acre-ft/year reduction for Jan1,2027–Dec31,2028. Not a current2026 deduction.

Source: IBWC Aug15,2025 and Sep9,2026 publications; Minutes323/330/334

Published initial allocation is not silently recomputed from rounded1.5million; treaty-volume/rounding reconciliation remains open. Morelos anchor is approximate regional geography, not surveyed engineering geometry.

Tribal rights and settlement context

decreed diversion, multi-source settlement budget, or reported consumption with its own recipient · Each instrument / listing has its own source date

Separate sovereign right, infrastructure access and actual use. Do not treat an unquantified right as zero. CA-side tribal facts retain CA-side scope.

GRIC653,500 acre-ft/year settlement budget includes CAP, Gila, Salt and groundwater. It is not all Colorado River consumption.

Source: Ten Tribes Partnership/Reclamation2018 Ch3; DOJ Gila settlement2023 update; Reclamation2026 entitlement listing and CY2025 accounting

Not a complete tribal rights register. Six Census reservation outlines are geographic context, not water-rights boundaries.

Critically overdrafted groundwater basins

DWR classification, not pumping/storage volume · DWR GIS boundary current Dec9,2022; classification2016 with2019 modifications

Use21 actual GIS polygons, simplified .004° with topology preservation; no interpolation from wells.

Kern County subbasin is selectable, with its DWR boundary date and long-term classification.

Source: DWR i08_CriticallyOverdraftedBasins critical_flag=Y

Not current groundwater level, live drought severity, or a modeled depletion quantity.

Rivers, conveyances, lake outlines and traces

nonquantitative physical / geographic context · Per-source geometry date

Source-backed lines with constant width; join exact endpoints only. Trace highlights documented relationships. Label unresolved Oroville outlet and Delta-channel routing breaks; do not invent a bridge.

Oroville trace includes Feather, Sacramento, Delta, Banks and SWP. Delta channels remain explicitly unresolved.

Source: USGS NHD, NaturalEarth, DWR SWP reaches, inherited OSM and MWD reference

Natural river source geometry is not engineering topology or measured end-to-end delivery. LVVWD remains EPA150-metre inventory-buffer approximation because operator API requires authentication. Unlabeled non-reservoir lakes are plain terrain water.

Where the water goes · M5 review

Census 2020 population controls label eligibility: at wide zoom at least 250,000, mid zoom 25,000, close zoom 10,000. Corrected Census 2020 urban areas are subtly shaded and labeled once; small places inside an urban area do not get another dot-and-name. These footprints are urban geography, not utility boundaries. GEOIDs preserve namesake identity. Community context cites operators and distinguishes groundwater replenishment from direct delivery.

Crop value estimates use the same crop, state and harvest year: statewide NASS gross production value divided by harvested acres, then by the 2023 Irrigation and Water Management Survey’s mean acre-feet applied per irrigated acre. Annual crop statistics and irrigation survey populations differ. This is a state-scale estimate, not a matched-farm measurement, profit, marginal value of water or consumptive use. Rice, potatoes and alfalfa hay are shown; suppressed cotton values remain omitted. Alfalfa is a qualified closest pairing: value and harvested acres cover dry hay and mixtures, while the water category also includes greenchop and silage. Each estimate labels this mismatch.

Delivery prices describe different services: IID canal delivery and MWD untreated or treated wholesale supply. Separate fixed charges can apply. They are context, not a farm-versus-city value ranking.

City water figures retain their own supplier geography, definitions and years. Metro GDP is BEA’s current-dollar, all-industry total, shown through FRED’s public mirror. The last official BEA metropolitan GDP year is 2023; BEA discontinued MSA estimates in February 2026. County GDP is not substituted. Metro GDP is never divided by water.

Powell contour playback

Playback uses daily Reclamation HDB elevation readings (NGVD29) and three published USGS Byerley 2024 outlines at 3,490, 3,520.46 and 3,525 feet NGVD29. It chooses the greatest available contour at or below each day’s elevation. These are coarse operational-tier shapes, not the full five-foot dataset or the actual shoreline. Missing readings remain gaps. Shapes are generalized without interpolation, scaling or morphing. Date, observed elevation and chosen contour stay visible. Playback jumps between real days when the chosen outline changes so the map visibly animates; the slider still exposes every daily position. Satellite polygons remain separately dated check frames. USGS extent polygons.

M6 reference outlines and exploration

All 16 reservoirs have USGS NHD high-resolution waterbody polygons, simplified without scaling. Their label anchors lie inside the reference polygon. These are source-vintage waterbodies, not current shorelines or operational full-pool contours. San Luis is offstream pumped storage; the other 15 have an intersecting source-river geometry. NHD source segments join the Gunnison and Colorado at Grand Junction, without an invented connecting line.

Reservoir polygons accept taps within 16 CSS pixels on desktop and 22 on touch. Reservoir labels have an additional 6-pixel hit buffer. Rivers and canals use 16 pixels on desktop, 20 on touch; points and reservoirs precede nearest-line resolution. Ordinary taps preserve the camera.

Layers → Elevation review options offers the existing paper fade, persistent hillshade and stronger contrast on the same retained relief raster. These are visual explorations, not new elevation data or a final design choice.