Why radar differs from your gauge

Same rain. Different ways of measuring it.

Your gauge catches rain at one point on the ground. Radar observes precipitation above the ground and estimates what fell across an area. It is reasonable for the totals to differ.

Radar looks above your garden

A radar beam rises with distance from the radar site. Mountains can partly block it, and precipitation aloft can change before it reaches the ground. Some rain evaporates on the way down. These effects make a radar estimate different from what lands in your gauge.

One point is not a whole area

NOAA’s live MRMS totals use roughly 1 km grid cells. A gauge samples a much smaller spot. A narrow shower can be heavy at your fence and lighter across the rest of the cell. Rounded coordinates are useful for finding a radar area; they do not make the result a measurement at your exact pin.

Wind and siting affect the gauge

Wind can carry drops over the gauge opening. Nearby trees, buildings and splashing can change what gets collected. A gauge is valuable local evidence, but its placement and reading routine matter.

Check the clock before the total

Inchfall’s ledger uses the 24 hours ending at 12:00 UTC on the report date, about 7 am Eastern in standard time. Your gauge may be read at a different local hour. A shower near the boundary can land on different dates in the two records.

The live “last 24 h (radar)” value is a rolling window and is separate from the morning ledger. Compare matching time windows before drawing conclusions.

Compare a storm, then a pattern

A multi-day total reduces some of the timing differences. In Inchfall’s limited November 2026 comparison of IEMRE’s radar-based estimates, multi-day totals had median errors of about 10–18% against CoCoRaHS volunteer gauges; wet days (at least 0.10 in) had daily median errors of about 25–35%; individual days can differ more. See the exact results and methods.

One or two readings are not enough to establish a stable local bias. Inchfall’s Free gauge log and side-by-side list provide context for one place; Pro adds the chart and optional correction. Optional correction is off by default and only unlocks after at least 10 wet readings totaling at least 2.0 in. It is not a promise of an exact match.

Keep the gaps visible

No data is different from no rain. Inchfall labels missing days instead of inventing a value. Recent values can be preliminary; waiting for the record to settle is often more useful than comparing a fresh estimate.

More about sources and quality labels →