Intertidal biofilm remote sensing paired spectra-biofilm training and validation data, South San Francisco Bay

Intertidal microbial biofilms, or microphytobenthos, support estuarine biogeochemical cycling, the physical stability of mudflats, and food webs, particularly those of migratory shorebirds. Photosynthetic biofilms dominated by diatoms, cyanobacteria and chlorophytes represent a significant fraction of biofilm biomass and contain pigments that can be detected with remote sensing. These diverse biofilm community types vary in indicator pigments and functional traits related to biogeochemical cycling and nutritional quality. We modeled and mapped the spatial variation in intertidal biofilm distribution, quantity, diversity, and functional traits using multi-scale spectroscopic data collected within southern San Francisco Bay, California, USA (South Bay). We developed single and multiple response partial least squares regression (PLS) models of chlorophyll-a (chl-a; biomass indicator), indicator pigments: fucoxanthin and diadinoxanthin (diatoms), zeaxanthin (cyanobacteria), and chl-b (chlorophytes), and functional traits: carbohydrates, lipids, and total organic carbon from paired in-situ biofilm data and field spectra collected with an ASD full-spectrum Field Spec Pro. Biofilm abundance, trait, and pigment data were obtained by analyzing biofilm samples in a lab. The PLS models were scaled to the South Bay with a 3.7m AVIRIS-NG (Airborne Visible/Infrared Imaging Spectrometer-Next Generation) hyperspectral image. The model %RMSE calculated from AVIRIS-NG test samples ranged from 12.7% for chl-a to 49% for chl-b; for six of the eight models, %RMSE was 23% or below. This data release contains .csv files for the paired ASD field spectra-biofilm calibration/validation dataset and the paired AVIRIS-NG-biofilm validation. This data release is associated with another data release that contains the published biofilm maps (Byrd et al. 2024, https://doi.org/10.5066/P13ZIMYC).

[
  "010:12"
]

[
  "geospatial"
]