Record-basis comparison
| Record | Meaning / units | Valid operation | What must not be inferred |
|---|---|---|---|
| Accumulated counter | Running depth, mm or inches | Known same-regime adjacent differences | First value is not rainfall during the first interval; decreases do not reveal reset-period rain |
| Interval depth | Depth over one stated interval | Sum known intervals; group under an explicit definition | A skipped/missing interval is not a zero |
| Interval intensity | Depth divided by interval hours | Compare with a matching accumulation duration | Not instantaneous peak or an IDF design value |
| Observed wet fragment | Known depths between dry or unknown boundaries | Describe its known depth/duration and boundary flags | Unknown boundaries do not establish an independent complete storm |
| Paired accumulation | Target/reference sums from common usable periods | Descriptive cumulative relation and selected segment ratios | Omitted pairs are not full-record rainfall; ratio change is not proof of cause |
| Fixed-network mean | Spatially weighted depth for a matched period | Use all required station observations and established weights | Missing stations do not become dry or donate their weight |
Time and missing-data contract
Counter/event CSV uses YYYY-MM-DD HH:mm in one supplied fixed-offset clock and an entered whole-minute step. Every time must be real, ordered and on the same grid. The tool does not infer UTC, daylight-saving changes or national date formats. Interval labels are ends, not starts. Paired/network period labels are preserved in supplied order; matching chronology and duration remain the operator's responsibility.
Blank and exact NA mean missing;0 means observed dry. Negative sentinels, inequalities, NaN, infinities, malformed rows and duplicate IDs are rejected, not silently omitted. Plain unquoted CSV keeps the bounded parser explicit. Preserve original quality flags separately.
Conversions that do not change the method
- 1 inch =25.4 mm.
- Interval intensity in mm/h = interval depth in mm ×60 / interval minutes.
- A segment ratio is unchanged by converting both station series with the same depth factor.
- A spatial mean changes depth units, not station weights.
- Source-unit selection does not convert pasted CSV values automatically.
Descriptive quantities, not acceptance rules
Accepted counter depth can be partial. Event count depends on a chosen dry gap and available boundary context. A double-mass ratio is undefined without usable positive reference accumulation. Available network weight describes observed coverage, not the unknown rainfall over the remainder. None supplies a universal calibration tolerance, event criterion or correction factor.
Why the same0 can mean different things
A known unchanged counter supports a dry increment; two zero station observations support a dry paired period; a complete all-zero network supports an areal mean 0. In contrast, an empty sum across no usable observations is not a measured rainfall total. The tools label those cases not observed or not assessed.
Export and precision
Screen/print tables show at most100 rows, with the limit stated. Audit CSV retains all analyzed output rows; only the counter's separate interval-input CSV is intended as direct input to the event tool. Copy retains configuration and full entered text. Ordinary browser floating-point arithmetic applies; supported nonzero inputs are1e-12 to 1e9. The double-mass chart can be thinned while calculations/export remain complete. No chart or preview is a substitute for the original instrument record.
Sources
- Colorado CDSS TSTool — Delta, missing endpoints and counter resets
- IETD method documentation — interval rainfall and chosen dry gaps
- Colorado CDSS TSTool — Cumulate and double-mass analysis
- NOAA NWS — Basic Hydrologic Concepts: mean areal precipitation
- NOAA NWS — Precipitation measurements and spatial methods