NASA IMERG analysis, early validation report

This dataset validation report plots a sample of values from the NASA IMERG analysis, early dataset over time and across space, comparing where possible to a previously validated reference dataset. It also reports the quantity of unavailable values and their associated timestamps. These analyses are one layer of a multi-layered dataset validation process we perform at dynamical.org and also provide users a preview of the dataset contents.

Report generation start time: 2026-07-24T03:55:31 UTC

Summary

This dataset has been reviewed and is ready for use. One note for users: the precipitation quality index reflects a change in the underlying satellite observing system partway through the record (see review notes below). Precipitation magnitudes themselves are unaffected.

What looks good

• A complete 30-minute record spanning the full IMERG era, 1998-01-01T00:00 through 2026-07-23T15:30 (500,672 timesteps), on the expected 0.1-degree global latitude/longitude grid.
• Geometry and units are correct: precipitation is reported as a rate in kg m-2 s-1 and the quality index is dimensionless (1).
• Precipitation is physically reasonable and agrees closely with the independent NOAA GEFS reference. At the 2026-07-22T12:00 snapshot the global spatial mean is 2.84e-05 kg m-2 s-1 for IMERG versus 3.02e-05 kg m-2 s-1 for GEFS, with matching spatial structure and a matching value distribution.
• No discontinuity is visible across the 2026-03 transition from IMERG version V07B to V07C.
• The quality index stays within its expected [0, 1] range throughout the record.
• In the recent GPM-era record, missing values are minimal — roughly 0.3-0.6% of grid cells at a given time, seasonally varying and concentrated toward the poles. No timestep anywhere in the archive is entirely empty (no all-NaN gaps). Coverage is lower and more variable earlier in the record (see review notes).

Review notes

• The precipitation quality index shows a shift in its typical level around 2014, when the source observing system transitioned from the TRMM era to the GPM Core Observatory era. This is a genuine change in the input measurement system, not a data error, and it does not affect the precipitation rate magnitudes.
• Spatial coverage is lower and more variable earlier in the record. Through roughly 2014 (the TRMM era, before the GPM Core Observatory) a larger share of grid cells are unmeasured at any given time — concentrated toward the poles, but also appearing as scattered gaps between satellite overpasses at lower latitudes — and coverage tightens from about 2014 onward. Precipitation values where present are unaffected.

Datasets

Run parameters

Availability

The plot below shows the availability of each variable over time — light red marks unavailable data, dark green marks present data.

• Unavailable timestamps: unavailable_timestamps.txt

Per-variable details

precipitation_quality_index_surface

Metadata

Point time series — 1998-01-01T00:00 - 2026-07-23T15:30

Spatial — 2026-07-22T12:00 (reference not available)

Temporal — 2020-03-03T22:59 - 2020-03-13T22:59

Availability — 500651 of 500672 positions complete (incomplete 1998-01-06T03:00 → 1998-01-06T13:00); nulls P1 0/500672, P2 21/500672

precipitation_surface

Metadata

Point time series — 1998-01-01T00:00 - 2026-07-23T15:30

Spatial — 2026-07-22T12:00 (reference at 2026-07-22T12:00)

Temporal — 2020-03-03T22:59 - 2020-03-13T22:59

Availability — 500651 of 500672 positions complete (incomplete 1998-01-06T03:00 → 1998-01-06T13:00); nulls P1 0/500672, P2 21/500672