Monday, January 13, 2020, 12:15PM - 1:15PM
Haruko Murakami Wainwright
Lawrence Berkeley National Laboratory
SEEC room S228 (Sievers Room)
Coffee and cookies at 11:45, as usual.
Microtopography – topographic variation on the scale of several meters – has a significant impact on snow distributions and hydrology, and hence on soil moisture and properties, biogeochemistry, microbial community and plant functional types and dynamics. Its profound impact has been observed even in ecosystems previously known as temperature-limited such as arctic tundra ecosystems. Recent advances in remote sensing and geophysics (including airborne LiDAR, UAV, electrical resistivity tomography) have enabled the characterization of microtopography and its impact on above-ground properties (e.g., snow, plant functional type, geomorphology) and below-ground properties (e.g., soil properties, soil moisture, permafrost) at sub-meter resolutions. Such high-resolution characterization opens a door to understand how microtopography influences ecosystem functioning, what is the scale at which topographic variability is important, and what is the impact of spatial aggregation and averaging, for example, to quantify the overall carbon budget. This talk first demonstrates multiscale multi-type data integration strategies for mapping the spatial heterogeneity of above/below-ground properties and carbon fluxes in two ecosystems: arctic ice-wedge polygonal tundra in Alaska and mountainous meadow ecosystems in the Rocky Mountains. We discuss (1) ecosystem zonation approaches based on unsupervised/supervised clustering, in which the landscape is classified into multiple zones associated with microtopographic features and larger geomorphic patterns, and (2) Bayesian hierarchical models for mapping key properties with uncertainty quantified. In particular, we demonstrate the use of Kalman filters to map the spatiotemporal variability of net ecosystem exchange (NEE) at the 0.5-meter resolution based on flux chamber measurements and airborne images. We then explore the effect of spatial aggregation on NEE in the Arctic tundra at different scales of heterogeneity. Results show that microtopography creates a significant heterogeneity in above/below-ground properties, and that microtopographic landscapes could create Simpson’s paradox; the situation in which a trend or bias appears or disappears depending on the level of spatial aggregation.
Free and open to the public.