Ecosystem-scale plant hydraulic traits estimated using model-data fusion

Abstract:

Droughts are expected to become more frequent and severe under climate change, increasing the need for accurate predictions of plant drought response. This response can vary considerably depending on plant properties that regulate water transport and storage within plants, i.e., plant hydraulic traits. Large-scale models usually parameterize plant traits based on plant functional types (PFTs). However, numerous studies have shown that plant hydraulic traits vary within PFT as much as across PFTs. Given that models incorporating plant hydraulics are becoming more common, it is essential to better map plant hydraulic traits. To do so, using in-situ measurements remains challenging given limited spatial coverage and high variability between and even within individual trees. Here, we use a model-data fusion approach to evaluate the spatial pattern of plant hydraulic traits across the continental US. This approach integrates a plant hydraulic model with microwave remote sensing products that inform ecosystem-scale plant water regulation. In particular, we use both surface soil moisture and vegetation optical depth (VOD) derived from the X-band JAXA Advanced Microwave Scanning Radiometer for EOS (AMSR-E). VOD is proportional to vegetation water content, which is here assumed to be a product of linear functions of leaf water potential and biomass. In addition, ET from the Atmosphere Land-Exchange Inverse model (ALEXI) is assimilated. The plant hydraulic traits (e.g. P50 describing xylem vulnerability and g1 describing optimal stomatal behavior) derived from remote sensing are compared with independent sources based on in situ measurements. We then use the resulting maps to investigate how environmental conditions shape the variation of plant hydraulic traits beyond PFTs. The results and the model-data fusion approach established here will facilitate parametrizing plant hydraulics in large-scale models for Improved predictions of drought response.