TOBS

TOBS, short for Time of Observation Bias, is a term used in meteorology and climatology to describe systematic biases that arise in temperature records when the time of day of observations changes or is inconsistent across stations and periods.

Because surface air temperature follows a diurnal cycle, recording at different times can shift the daily mean, as well as reported daily maximum and minimum temperatures. Even small shifts in observation time can produce detectable biases in long-term trends if not accounted for, and these biases can vary by region and season.

TOBS corrections are applied during data homogenization to align daily temperature series to a common reference observation time. Methods rely on station metadata, overlap periods, or statistical detection of shifts to estimate and remove the bias. Large climate datasets and reanalysis products, including the Global Historical Climatology Network (GHCN) and CRUTEM, incorporate TOBS corrections to improve the reliability of long-term trends.

In some data formats, the field labeled tobs denotes the observed temperature value itself, distinguishing it from model outputs or reanalysis estimates.