(2019) Nonstationary frequency pairing reveals a highly sensitive peak flow regime to harvesting across a wide range of return periods

By XuJian Joe Yu and Younes Alila

A common approach used by forest hydrologists to understand and quantify how forest harvesting affects floods was proclaimed flawed by a 2009 landmark study. With a century of forest hydrology research now questioned by academics and by governments alike, Yu and Alila were compelled to develop a new method, known as nonstationary frequency analysis, to evaluate continuous forest harvesting and regrowth effects on floods. The application of this method is demonstrated on the Greata Creek Camp paired watershed study site in the snow environment of British Columbia, Canada. The method reveals a highly sensitive flood regime to forest harvesting in a mid-elevation south-facing zone of the watershed, summarized as follows:

1. Tree removal increased the magnitude and frequency of floods across not only small and medium (return periods < 10-yr) but also the large events (return periods > 10-yr).
2. Floods can be sensitive to even small rates of forest harvesting, depending on location within the watershed. For instance, the removal of only 11% of the watershed’s trees doubled the frequency of floods of all sizes, as a result of a 9 –14% increase in the magnitude of such floods.
3. Depending on the extent of harvested area and the size of flood events, larger harvest or cut rates resulted in two, three, and fourfold increases in the frequency of large floods.

These outcomes run counter to the prevailing wisdom in hydrological science. The co-authors suggest that these outcomes have major implications for the safety of drainage structures such as bridges and dams, human settlements, drinking water quality, and the sustainability of riverine ecosystems. The co-authors call for a re-evaluation of past studies based on this new method.
For further information, contact Younes Alila at younes.alila@ubc.ca

Read full study