How Can Sampling Frequency Change a Temperature Summary?
From the earlier reference collection. Read the current software blog.
Sparse or irregular samples can miss short-lived changes and make averages or daily extremes describe only the observed points.
Samples do not cover every moment
A sensor summary is based on observations that were actually collected, not on every instant in the period. If the sensor samples once per hour, a brief warm peak between samples may be missed. If it samples every minute, it has more chances to capture short changes, though the instrument and placement still shape what it records.
The observation schedule should accompany a summary. A daily maximum calculated from hourly samples is the highest recorded hourly sample, not necessarily the highest temperature that occurred that day. Using precise wording keeps readers from assigning more coverage to the data than the sampling provides.
Irregular intervals affect averages
A simple mean of the available values gives each observation equal weight. When samples are evenly spaced, that may be a reasonable summary of the observed period. When intervals vary, equal weighting can overemphasize clusters of samples and underrepresent long gaps.
Imagine readings taken at 09:00, 09:05, and 15:00. An unweighted mean gives the three observations equal influence even though two are close together and one follows a long interval. A time-weighted approach may be more appropriate for some questions, but its assumptions and method should be stated. Neither calculation fills in unknown conditions automatically.
Daily extrema depend on coverage
A reported minimum or maximum is limited by how often and when samples were collected. A sensor that misses overnight periods may not capture the daily minimum. A long data gap during the afternoon could conceal the maximum. Record the number of samples or coverage where practical, and mark days with substantial gaps.
For example, “daily maximum among 22 reported hourly observations” is clearer than “daily maximum” when two expected hours are absent. If a system estimates missing periods, distinguish the estimated extreme from one directly observed. The distinction matters when readers compare summaries across days or stations with different reporting schedules.
Compare like with like
Before comparing averages or extremes from two sensors, check their sampling intervals, missing periods, time zones, and aggregation methods. A station sampled every five minutes may capture short changes that an hourly station misses. That difference can make its recorded maximum higher without indicating a faulty instrument.
A useful report states the summary rule and the observations available to it. Keep original timestamps so users can recalculate the statistic if needed. When coverage is too sparse to support the intended claim, show the limitation instead of presenting a number with unwarranted confidence.