Data relating to non-flood years is presented on the Single Site dashboard, the Growth Curve pooling group table and the QMED Donor adjustment table. The Growth Curve pooling group and the QMED Donor adjustment tables also contain information on trends in the AM data. Non-flood year data is calculated using the AM data within WINFAP, whereas the trend statistics are extracted from the NRFA Peak Flow Dataset.

Non-flood years

The number and percentage of non-flood years (identified as those years in which the annual maxima is less than half the QMED) is presented on the Single Site dashboard, the Growth Curve pooling group table and the QMED Donor adjustment table. Non-flood years can have a significant impact on the L-moments used in the statistical method hence it can be important to adjust for the presence of these.

The methods for identifying non-flood years, and adjusting the L-moments to account for these, are described in Flood Estimation Handbook Volume 3, Chapter 19. This adjustment was previously known as the 'Permeable Adjustment'. Within the report 'Review of flood frequency estimation in groundwater dominated catchments' (JBA, 2022) it was noted that catchments with a high proportion of non-flood years tend to be those with low rainfall (SAAR) and are not restricted to permeable catchments hence this is now more widely applied.

Presentation of non-flood year information in the relevant tables allows you to identify whether to apply the non-flood years adjustment. If an adjustment is required this can be completed using the WHS Non-Flood Years Adjustment Spreadsheet. Please see regulators guidance for more details of how and when this should be applied.

As the QMED is the median of the AM values, non-flood years will rarely have an impact on this value hence adjustment to account for non-flood years is not generally necessary.

Trends

Trend statistics, in the form of the Mann-Kendall Z statistic (MKZ) and the related significance, are provided in the NRFA Peak Flow Dataset from version 12 onwards and are updated annually. More details can be found on the NRFA trend pages. The trend information is presented in the Growth Curve pooling group and QMED Donor adjustment tables. As these are not calculated by WINFAP, these will NOT be updated if the AM data is edited.

A dynamic link to the NRFA Flow Trend Explorer AMAX plots for the specific station is provided in the Growth Curve pooling group table and the QMED Donor adjustment table. This is a web based app that uses NRFA data to allow users to visualise trends in the NRFA dataset. Viewing the 'Triangle Trends' plots can also provide useful information on non-stationarity at a station.

A negative value of MKZ denotes a negative (decreasing values) trend and a positive value a positive (increasing values) trend. Where the strength of the trend is statistically significant at the 5% or 10% level (where 5% is of greater confidence) this is specified in the table.

Where there is a trend in a dataset, this is often referred to as having 'non-stationarity'; that is the underlying distribution of AM data may change over time. As the assumption of the statistical method is that the underlying data represents a stationary period of record this can have an impact on the estimation of flood peak flows.

Whilst a trend may be present due to climate change, there are many other reasons (e.g. hydrometry, land use change) why a trend may be present. Identifying and allowing for non-stationarity in flood estimation can therefore be very complex and these values are provided in WINFAP for information only. Please see regulators guidance for more details of how and when trends or non-stationarity should be taken into account.