RainReserveFIELD NOTES / WATER · 001How it works ↗

SMALL SYSTEMS, CLEARER CHOICES

Make room
for the rain.

A bigger barrel isn't always a better answer. Explore what your roof could collect, what your garden could use, and what still spills over.

Try a water-balance scenario ↓

A transparent learning tool. Not a weather forecast or installation plan.

ONE DROP, THREE POSSIBILITIES

Catch it. Keep it.
Put it to work.

Trace the flow with hover or keyboard focus.
ILLUSTRATION · NOT LIVE DATA

01 / THE EXPLORER

Follow every litre.

SYNTHETIC DATA · 30 DAYS

Your setup

Invented daily rainfall, not local observations. All scenarios begin with an empty barrel.

Use your own rainfall

Compare the same barrels with daily rainfall you supply. Nothing is uploaded or saved.

UTF-8 CSV, up to 64 KiB: date,rain_mm, then 1–366 consecutive days, oldest first. Dates: YYYY-MM-DD. Daily rain: 0–500 mm. No missing days; use 0 only when you know no rain fell.

The example is synthetic, not measured weather. Your values are not independently verified.

Targets are assumptions, not plant-care advice. Direct rain reduces that day's target; soil storage is not modeled.

0 L2,000 L

Change assumptions, then update. The capacity slider updates immediately when inputs are valid.

SCENARIO / PERIOD TOTAL

— L used from storage

AT THE END OF DAY 1
in storage
Collected
Used
Overflow
Unmet target

Your rainfall, day by day

End storage Rain

Synthetic example · 30 days

DAY 1DAY 30

Bars show storage relative to capacity. Dots mark rainy days. Select any day for daily values, rounded to 0.1 L.

02 / THE TRADE-OFF

More storage.
What actually changes?

30-day balanceBaselineScenario

Take the assumptions and every daily balance with you.
CSV export stays on your device. No account, tracking, or API calls.

03 / UNDER THE SURFACE

A model you
can question.

Useful comparisons start with visible assumptions. These results are hypothetical, not measured water savings or environmental impact.

The daily calculation

1 mm falling on 1 m² equals 1 litre. Collected water = rainfall × connected roof area × collection efficiency. Rain enters storage first; anything above capacity overflows. Then the barrel supplies the remaining daily garden target.

Garden target = max(0, daily water target − rainfall) × garden area. The same rainfall is applied to roof and garden. End storage = start storage + collected − overflow − used. Unmet target = garden target − used.

What this leaves out

No soil moisture carried between days, evaporation, first-flush diversion, leakage, seasonal plant needs, irrigation losses, hourly storm timing, water quality, costs, or local constraints. Same-day rainfall is treated as fully effective on the garden. These simplifications can overstate useful supply. An empty start keeps stored water provenance clear.

Capacity and collection efficiency are user assumptions. Input limits keep the explorer bounded; they are not installation recommendations. This is not guidance for drinking water, plumbing, structural safety, or legal compliance.

Why explore rainwater storage?

EPA WaterSense describes rain barrels as one way to collect rainfall for outdoor use, alongside reducing overwatering. Actual usefulness depends on local conditions and demand. This explorer helps compare capacity trade-offs before seeking site-specific advice.

EPA WaterSense: Watering Tips ↗ · Guidance checked October 28, 2026.

Data, privacy & authorship

All four rainfall patterns are hand-authored synthetic examples. None represent a city, measurement, or forecast. Everything runs locally in your browser. Nothing is saved automatically; refreshing resets the demo. CSV downloads include settings, rainfall provenance, and both daily balances. You may also import your own consecutive daily rainfall in millimetres. Imported dates and values are supplied by you, not verified observations or forecasts; the calculated balances remain hypothetical. Files are read only in this browser and are cleared by Reset demo or refresh. Missing rainfall is never filled in automatically.

Original prototype by Sharon Basovich with Codex AI assistance, October 28, 2026. No third-party visual assets or runtime dependencies.