Trends in Annual Maxima

Select the station to explore the AMAX trends of. Return periods are in years (QMED is the 2-year event). Drag on the plot to select a time-period; the graph will automatically recompute trends for the highlighted period.
Hover over the graph to read off values. Click the zoom button to switch to zoom mode, and click 'box select' to return to data selection. For the Mann-Kendall standardised score (MKZs) an asterisk * indicates significance at 5% level.

Catchment info

Trends in Daily Flow

Select the station to explore the daily mean flow trends of. Exceedences are percentiles of gauged daily flow, and are computed on a yearly resolution. Drag on the plot to select a time-period; the graph will automatically recompute trends for the highlighted period.

MKZ test statistics for daily flow:

Hover over the graph to read off values. Click the zoom button to switch to zoom mode, and click 'box select' to return to data selection. For the Mann-Kendall standardised score (MKZs) an asterisk * indicates significance at 5% level.

Catchment info

Multi-temporal Trend Summary

Select the station to explore the AMAX multi-temporal trend analysis. Adjust the slider to see different periods. Only stations with more than 26 years of data and which have AMAX data are included.
Each grid square shows the trend observed if the data is restricted to a specific time period. Red indicates negative trends, blue indicates positive. Green points highlight significant trends.

About this Shiny App

This Shiny app provides a way of exploring and evaluating trends in peak flow and daily flow data within the measuring stations with records in the National River Flow Archive. The main purpose is to highlight trends in peak flow and daily flow percentiles by using the Theil Sen estimator of trend.

The outputs here are based on data taken using the NRFA API, and are designed to match the station trends pages on the NRFA webpage.

How the app works

On each tab, you can select a station and the data will be plotted along with lines of trend or long-term stationary averages. On each plot, you can highlight a section of the data and the trend will be shown based on just the highlighted section, showing the sensitivity of trend calculation to length of record.

On the "Triangle Trends" page, a full summary of the AMAX analysis is presented for those stations with more than 26 years of data, where all the possible periods of record are shown, and the associated trends.

Note that only stations of sufficient quality are included for AMAX analysis.

Fitting Trends

The trends on daily flow and peak flow are based on the Theil-Sen estimator (Sen 1969), and the Mann-Kendall test is used to test significance, as in (Harrigan et al., 2018). Return periods of peak flow events are fitted using a Generalised Logistic distribution, and a linear trend is included in the location parameter using the Theil-Sen estimator. Nonstationary return periods are calculated by evaluating the nonstatinary parameters in each year of interest and computing the relevant quantiles, as in (Griffin et al., 2019). Here trend is only computed for average behaviour, so the same trend is presented for all rarities of flood.

Funding

This work was supported by the Natural Environment Research Council award number NE/R016429/1 as part of the UK-SCAPE programme delivering National Capability.

UKCEH 2020