Barcelona Temperature and Precipitation Tracker

Daily temperature displays either the original daily values or a 7- or 15-day trailing mean. A trailing mean averages the plotted date and the preceding 6 or 14 days. It is shown only when the plotted date itself has an observation and at least 70% of the selected window contains valid data.

For daily temperature, the selected historical reference defines the grey distribution. For every calendar date, the chart calculates the median and the 10th–90th percentile range across the selected historical years after applying the chosen smoothing. Missing values are excluded from that date's calculation. For daily temperature comparisons, 29 February has its own calendar position and reference values are calculated from leap years only.

Daily temperature anomaly is the temperature for a date minus the mean for that same calendar date during the selected 30-year baseline. The zero line therefore represents the selected baseline climatology.

Cumulative year-to-date mean temperature anomaly is the mean of all available daily anomalies from 1 January through the plotted date. It is already cumulative, so no additional smoothing is applied. The displayed ranking compares the latest year with years having at least 90% valid coverage through the same date.

Daily precipitation is shown as bars for the selected year, together with the cumulative precipitation total from 1 January. The selected historical reference defines the cumulative precipitation median and 10th–90th percentile envelope for each calendar date, and its arithmetic mean is used for the displayed percentage of normal. The year-to-date rank and percentile compare cumulative precipitation through the same calendar date across all years with complete precipitation coverage from 1 January through that date; the percentile is the percentage of complete historical YTD totals at or below the selected value. Historical years are aligned by calendar date rather than numerical day of year; on 29 February, the displayed historical median and percentile envelope are linearly interpolated between 28 February and 1 March so that leap and non-leap years remain comparable. Daily precipitation itself is not ranked or assigned a percentile.

Monthly, seasonal and annual statistics use completed calendar periods. Temperature means require at least 90% daily coverage; precipitation totals require complete daily coverage. Anomalies are relative to the arithmetic mean for the same period during the selected 30-year climatological baseline. Temperature standardized anomalies divide the temperature anomaly by the sample standard deviation across complete baseline values. Because precipitation totals are strongly skewed, standardized precipitation anomalies instead use an empirical normal-score transformation of the selected 30-year baseline, with negative values indicating drier-than-normal and positive values wetter-than-normal conditions. Precipitation rankings are reported from whichever extreme is closer (for example, 3rd driest rather than 99th wettest), while the percentile always runs from dry to wet. Historical ranks use all complete values in the Fabra record.

The monthly climate heatmap can display temperature means, temperature anomalies, standardized temperature anomalies, precipitation totals, precipitation anomalies, or standardized precipitation anomalies. Temperature months require at least 90% daily coverage, whereas precipitation months require complete daily coverage. Anomaly and standardized-anomaly views use the selected 30-year reference period. For raw monthly values, tooltips also report the historical rank and percentile among all complete values for the same calendar month; precipitation ranks use whichever extreme is closer (driest or wettest). Rankings are not shown for the 11-year trailing-mean view. The 11-year view averages the selected monthly metric over the 11-year period ending in each plotted year. Missing or incomplete months are not interpolated.

Long-term trends are calculated separately for annual and seasonal temperature means and precipitation totals. Temperature periods require at least 90% daily coverage; precipitation periods require complete daily coverage. Linear trends are estimated by ordinary least squares, with Newey–West heteroskedasticity- and autocorrelation-consistent uncertainty using Bartlett weights and an automatically selected lag. Slopes are reported per decade together with 95% confidence intervals, p-values and R².

All curves and period statistics represent this single, elevated observation site; they are not spatial averages and should not be interpreted as temperature or precipitation averaged across the Barcelona metropolitan area.