Data from: Predicting photosynthesis-irradiance relationships from satellite remote-sensing observations
Abstract
Photosynthesis-irradiance (PI) relationships are important for phytoplankton ecology and for quantifying carbon fixation rates in the environment. However, the parameters of PI relationships are typically unknown across space and time. Here we use machine learning, satellite remote-sensing, and a database of in-situ PI relationships to build models that predict the seasonal cycle of PI parameters as a function of satellite-observed variables. Using only surface light, temperature, and chlorophyll, we achieve an R2 of 58% for predicting photosynthesis rates at saturating light and an R2 of 78% for predicting the light saturation parameter. Predictability is maximized when averaging environmental covariates over 30-day and 25-day timescales, respectively, indicating that environmental history and community turnover timescales are important for predicting in-situ PI relationships. These results will help improve the parameterization of satellite-based primary production models and quantify emergent environmental integration timescales in photosynthetic communities.
Dataset DOI: 10.5061/dryad.w6m905r1j
Description of the data and file structure
Data are a subset of the database curated by Bouman et al. (2018; Photosynthesis-irradiance parameters of phytoplankton: synthesis of a global data set. Earth System Science Data 10: 251–266) and updated by Kulk et al. (2020; Primary production, an index of climate change in the ocean: Satellite-based estimates over two decades. Remote Sensing 12: 1–26). All measurements analyzed here were taken using 14C uptake as a measure of photosynthetic rate using incubations between 1.5-4 hours in duration. The database reports parameters for fitted PI relationships. We chose subsets of PI relationships originating from established regional sampling programs wherein the same locations are sampled across time and over a range of environmental conditions using consistent sampling and experimental methodology. Analyzing these subsets acts to minimize bias when collating datapoints that differ in collection methodology. Samples were taken from the mixed layer based on an established monthly mixed layer depth climatology to best correspond with satellite observations. We matched PI relationships with satellite-derived photosynthetically available radiation (PAR), sea surface temperature (SST), and chlorophyll-a (Chl). We use 4km daily-averaged surface PAR from merged MODIS and SeaWIFS version R2018.0. SST at 6km resolution is from version 2.0 of the multi-sensor Operational Sea Surface Temperature and Ice Analysis (OSTIA) foundation temperature product. We use Chl from version 5.0 of the multi-sensor Ocean Colour Climate Change Initiative (OC-CCI) 4km product. We also used SST and Chl to compute an estimate of the picophytoplankton contribution to Chl, denoted according to equations given in Brewin et al. (2010; A three-component model of phytoplankton size class for the Atlantic Ocean. Ecological Modelling 221: 1472–1483). We constructed a series of temporal averages of the variables, with each average extending one day further back in time. We do this over 100 days for individual in-situ sampling locations, generating 100 covariate datasets to be used in the modeling of the PI parameters as a function of environmental integration timescale.
Files and variables
File: dat.csv
Description:
Variables
- sst: sea surface temperature [degree Celsius]
- chl: surface chlorophyll-a concentration [mg Chl m^-3^ ]
- par: surface incident photosynthetically active radiation [Einstein m^-2^ day^-1^]
- parml: mixed-layer depth averaged PAR [Einstein m^-2^ day^-1^]
- nano_pico: concentration ratio of nano-phytoplankton to pico-phytoplankton [unitless]
- micro_nano: concentration ratio of micro-phytoplankton to nano-phytoplankton [unitless]
- pico: proportion of phytoplankton represented by pico-phytoplankton [proportion]
- micro_pico: concentration ratio of micro-phytoplankton to pico-phytoplankton [unitless]
- lat: latitude of the PI experiment [degrees]
- lon: longitude of the PI experiment [degrees]
- depth: depth of water sampled for PI experiment [m]
- month: month in which the PI experiment was performed [number]
- PBmax: measured PBmax from the PI experiment [mg C mg Chl^-1^ hour^-1^]
- alpha: measured alpha from the PI experiment [mg C mg Chl-1 seconds-1 (micromol photons m^-2^ seconds^-1^)^-1^]
- Ek: ratio of measured PBmax and alpha [micromol photons m^-2^ seconds^-1^]
- region: sub-region used in the analysis [ID]
- date: date of the PI experiment [month/day/year]
- mld: climatological mixed layer depth according to the latitude, longitude and month of the sampling location [m]
Code/software
All code used to perform analyses and produce figures will be publicly available at: https://github.com/gregbritten/pi_parameters_public
Access information
All data used in this study are cited in the main text and publicly available with the following DOIs: OC-CCI chlorophyll (doi:10.5285/5011d22aae5a4671b0cbc7d05c56c4f0); OSTIA SST (doi:10.48670/moi-00168); MODIS and SeaWifs PAR (doi:10.5067/AQUA/MODIS/L3B/PAR/2022.0, doi:10.5067/ORBVIEW-2/SeaWiFS/L3B/PAR/2022.0).