Determining the suitability of remotely sensed snow disappearance date as a proxy for the onset of stem radial growth in conifers at the forest-tundra ecotone

Abstract:

Past research has shown that the snow disappearance date (SDD) corresponds closely to the onset of tree stem growth in the boreal forest which is an important component of the terrestrial carbon cycle. Many remote sensing platforms, such as the Moderate Resolution Imaging Spectroradiometer (MODIS) aboard the Aqua and Terra satellites, can detect SDD and could thus provide an approach for determining the onset of the growing season in high-latitude forests. The main goal of this study was to determine if remotely sensed SDD could be used as a proxy for determining the onset of stem radial growth at the forest-tundra ecotone (FTE). The SDD was obtained using two different approaches: when individual tree level measurements of soil temperature increased above 0^oC and MODIS derived SDD. These dates were compared to the date of stem radial growth onset derived from high-resolution point dendrometers affixed to each study tree. Data were collected over two growing seasons (2018 and 2019) and from two different locations at the FTE: the Brooks Range, Alaska, USA, and near Inuvik, Northwest Territories, Canada. Mean SDD occurred earlier in the Brooks Range than at the Canada site. Preliminary results also suggest that MODIS SDD occurred earlier than soil temperature derived SDD, and at both sites the SDD derived from MODIS and soil temperature was significantly different than the date of stem radial growth onset (p < 0.05). In Alaska, soil temperature derived mean SDD occurred approximately 10 days earlier than the mean onset of stem radial growth, whereas in Canada the mean onset of stem radial growth occurred 21 days earlier than mean SDD. These results suggest that SDD does not correspond with the start of stem radial growth in conifers at the FTE. This uncoupling between soil temperatures, SDD, and tree growth is somewhat surprising and demonstrates that the onset of stem radial growth is perhaps driven by biophysical factors not tested in this study, such as increasing air temperatures or photoperiod. Based on these results it can be concluded that remotely sensed SDD would not be an effective proxy for determining the onset of stem radial growth.