Large Wood dynamics during the Vaia storm : insights from post-event remote and field-based data

Forested basins undergo constant reshaping and self-regulation, especially during large disturbances such as windstorm events, which can have a compound impact on water, sediment, and forests. These events may lead to the recruitment of Large Wood (LW) from various sources, further influencing the morphology and dynamics of rivers. This study focuses on the Tegnas Torrent, located in the northeastern region of Italy (Dolomites, Belluno Province), featuring a drainage area of 52 km2. The main channel is fed by two significant tributaries, the Bordina and Angheraz streams, and spans a length of 9.5 km with an average slope of 3% and a predominant braided morphology. The Tegnas Torrent experienced severe consequences due to the Vaia storm (October 27th- 31st 2018). This event caused widespread tree windthrows, intense precipitation, and significant fluvial changes, including lateral migrations, erosions, and LW recruitment. The peak discharge exceeded 150 m3 s-1 (> 100-year recurrence interval). This study aims at assessing LW dynamics by exploiting post-flood remote sensing and field-based data to quantify the streambank recruitment of LW and the load. The LW recruited was quantified by calculating the volume of wood within the area eroded during the event (previously identified by a Dem of Difference analysis featuring the before and after event DTMs). To obtain the trees’ diameter at breast height (dbh), hence the volume, correlation between the Canopy Height Model (CHM) obtained by LiDAR and the height measured in 11 field-sampling plots were used. In other words, the post-event CHM was used as proxy to transfer the hypsometric curve (height ~ dbh) found in the field. The Tegnas Torrent's LW load, instead, was quantified through a field campaign performed in 2019. Nine 100m-long transects were surveyed. Logs longer than 1 m and with a diameter over 0.1 m were georeferenced and measured for length and diameter. The LW load representative for the whole channel (m3 ha-1) was then obtained, firstly, by averaging the load measured inside each transect and, secondly, by multiplying the total load by the area of the active channel post-Vaia (28.76 ha). The total volume of the recruited LW was 132 m³ ha⁻ ¹ while the overall LW load was 18 m³ ha ⁻¹ (496 m³). As far as the latter is concern, 466 LW elements were measured, ranging from 1 m to 28 m in length and 0.10 m to 0.90 m in diameter. The mean diameter and length were 0.18 m and 4.62 m, respectively, with a mean volume of 0.20 m³. Therefore, of the total recruited volume, 76% was transported outside the area of interest, while 24% was stored inside the enlarged active channel. Overall, the Tegnas Torrent demonstrated a LW load of 18 m³ ha⁻¹, consistent with local and global values for mountainous forested basins that were detected following low to high magnitude events. However, the Tegnas Torrent stands out in terms of LW recruitment, with a notable value of 132 m³ ha⁻¹. This highlights the significant impact of extreme events on large wood recruitment, offering crucial insights into general fluvial responses to disturbances. Nonetheless, given the spatial variability of LW recruitment volumes along area eroded, this study highlights the importance of further analyses concerning hydrological modelling validation for simulating the event and, thus for predicting streambank LW recruitment and mitigating risks during future scenarios.