Downscaled climate projections of future mesopelagic habitat in the California Current Ecosystem
Data files
Aug 11, 2025 version files 254.83 MB
-
light.csv
10.72 KB
-
meso_hab.csv
4.67 KB
-
mesohabitat_proj_2000-2100.nc
254.72 MB
-
n2_no3_chl.csv
13.04 KB
-
oxygen.csv
13.02 KB
-
Projections_analysis_R.Rmd
39.22 KB
-
README.md
13.82 KB
-
temp_iso.csv
8.24 KB
Abstract
Although the mesopelagic zone occupies a substantial volume of the world's oceans, our results suggest that the livable portion may compress vertically by ~40 m or ~39% by the end of the century. Using an ensemble of three downscaled climate projections from a high emissions scenario, we evaluated the connection between anthropogenic greenhouse gas emissions and changes in light and oxygen at depth, which influence the upper and lower limits of mesopelagic habitat in the central California Current. Although the model projects a small deepening (~ 2 m) of the upper light boundary consistent with increased stratification and reduced upper ocean productivity, the main driver of vertical mesopelagic habitat compression is the significant shoaling (by ~44 m) of the hypoxic boundary over the course of the 21st century. Differences in dissolved oxygen across ensemble members highlight the potential influence of equatorial dynamics and the California Undercurrent in constraining the future availability of mesopelagic habitat along the U.S. West Coast. Mesopelagic ecosystems connect the surface ocean to the deep sea, and a projected decrease in the vertical extent of mesopelagic habitat could have cascading effects on a broader range of marine ecosystem processes and carbon export.
https://doi.org/10.5061/dryad.kh18932hn
Description of the data and file structure
The following oceanographic data and R code accompany the manuscript "Projected 21st century compression of mesopelagic habitat in the California Current" submitted to Scientific Reports. Using an ensemble of three high-resolution downscaled climate projections for the California Current under a high emissions scenario, we evaluated potential changes in the availability of mesopelagic habitat by the end of the 21st century. The upper and lower boundaries of the mesopelagic zone are defined as the depth at which light intensity reaches (0.0217 W/m2) and the depth of the hypoxic boundary defined as ~63 mmol/m3. Additional ocean metrics include: temperature at 150 m, dissolved oxygen at 300 m, Integrated chlorophyll in the upper 100 m, maximum buoyancy frequency (N^2) in upper 100 m, light intensity at 150 m and mean nitrate (NO3) in the upper 100m.
Files and variables
File: light.csv
Description: Downscaled climate projections (2000-2100) of light metrics averaged over our study domain (see Figure 1 in manuscript). Ensemble mean and spread were calculated as the mean and standard deviation (sd) of the 3 projections from ROMS-GFDL, ROMS-IPSL and ROMS-HADL (see manuscript for full descriptions).
Variables
- LIGHT_150_ENSEMBLE: Ensemble mean of light intensity (units of W/m^2) at a depth of 150 m.
- LIGHT_150_ENSEMBLE_SD: Standard deviation (ensemble spread) of LIGHT_150_ENSEMBLE
- LIGHT_150_GFDL: Spatial mean of light intensity at 150 m (2000-2100) from ROMS_GFDL
- LIGHT_150_IPSL: Spatial mean of light intensity at 150 m (2000-2100) from ROMS_IPSL
- LIGHT_150_HAD: Spatial mean of light intensity at 150 m (2000-2100) from ROMS_HAD
- EUPHOTIC_DEPTH_ENSEMBLE: Ensemble mean of euphotic depth (light ~1% of surface) (units of m)
- EUPHOTIC_DEPTH_ENSEMBLE_SD: Standard deviation (ensemble spread) of EUPHOTIC_DEPTH_ENSEMBLE
- EUPHOTIC_DEPTH_GFDL: Spatial mean of euphotic depth (2000-2100) from ROMS_GFDL
- EUPHOTIC_DEPTH_IPSL: Spatial mean of euphotic depth (2000-2100) from ROMS_IPSL
- EUPHOTIC_DEPTH_HAD: Spatial mean of euphotic depth (2000-2100) from ROMS_HAD
- DEPTH_LIGHT_REF_ENSEMBLE: Ensemble mean of the depth at which light reaches 0.0217 W/m2 (units of meters)
- DEPTH_LIGHT_REF_ENSEMBLE_SD: Standard deviation (ensemble spread) of DEPTH_LIGHT_REF_ENSEMBLE
- DEPTH_LIGHT_REF_GFDL: Spatial mean of reference light level depth (2000-2100) from ROMS_GFDL
- DEPTH_LIGHT_REF_IPSL: Spatial mean of reference light level depth (2000-2100) from ROMS_IPSL
- DEPTH_LIGHT_REF_HAD: Spatial mean of reference light level depth (2000-2100) from ROMS_HAD
File: n2_no3_chl.csv
Description: Downscaled climate projections (2000-2100) of oceanographic metrics averaged over our study domain (see Figure 1 in manuscript). Ensemble mean and spread were calculated as the mean and standard deviation (sd) of the 3 projections from ROMS-GFDL, ROMS-IPSL and ROMS-HADL (see manuscript for full descriptions).
Variables
- N2_ENSEMBLE: Ensemble mean of max buoyancy frequency in upper 100 m (s^-1)
- N2_ENSEMBLE_SD: Standard deviation (ensemble spread) of N2_ENSEMBLE
- N2_GFDL: Spatial mean of max buoyancy frequency in upper 100 m (2000-2100) from ROMS_GFDL
- N2_IPSL: Spatial mean of max buoyancy frequency in upper 100 m (2000-2100) from ROMS_IPSL
- N2_HAD: Spatial mean of max buoyancy frequency in upper 100 m (2000-2100) from ROMS_HAD
- CHL_100_ENSEMBLE: Ensemble mean of integrated chlorophyll in the upper 100 m (mg/m^3)
- CHL_100_ENSEMBLE_SD: Standard deviation (ensemble spread) of CHL_100_ENSEMBLE
- CHL_100_GFDL: Spatial mean of integrated chlorophyll in the upper 100 m (2000-2100) from ROMS_GFDL
- CHL_100_IPSL: Spatial mean of integrated chlorophyll in the upper 100 m (2000-2100) from ROMS_IPSL
- CHL_100_HAD: Spatial mean of integrated chlorophyll in the upper 100 m (2000-2100) from ROMS_HAD
- NO3_100_ENSEMBLE: Ensemble mean of mean nitrate in the upper 100 m (mmol/m^3)
- NO3_100_ENSEMBLE_SD: Standard deviation (ensemble spread) of NO3_100_ENSEMBLE
- NO3_100_GFDL: Spatial mean of mean nitrate in the upper 100 m (2000-2100) from ROMS_GFDL
- NO3_100_IPSL: Spatial mean of mean nitrate in the upper 100 m (2000-2100) from ROMS_IPSL
- NO3_100_HAD: Spatial mean of mean nitrate in the upper 100 m (2000-2100) from ROMS_HAD
File: oxygen.csv
Description: Downscaled climate projections (2000-2100) of oxygen metrics averaged over our study domain (see Figure 1 in manuscript). Ensemble mean and spread were calculated as the mean and standard deviation (sd) of the 3 projections from ROMS-GFDL, ROMS-IPSL and ROMS-HADL (see manuscript for full descriptions).
Variables
- DEPTH_OX_63_ENSEMBLE: Ensemble mean of hypoxic depth (63 mmol/m^3) (units of m)
- DEPTH_OX_63_ENSEMBLE_SD: Standard deviation (ensemble spread) of DEPTH_OX_63_ENSEMBLE
- DEPTH_OX_63_GFDL: Spatial mean of hypoxic depth (2000-2100) from ROMS_GFDL
- DEPTH_OX_63_IPSL: Spatial mean of hypoxic depth (2000-2100) from ROMS_IPSL
- DEPTH_OX_63_HAD: Spatial mean of hypoxic depth (2000-2100) from ROMS_HAD
- OX_100_ENSEMBLE: Ensemble mean of dissolved oxygen at 100 m depth (units of mmol/m^3)
- OX_100_ENSEMBLE_SD: Standard deviation (ensemble spread) of OX_100_ENSEMBLE
- OX_200_ENSEMBLE: Ensemble mean of dissolved oxygen at 200 m depth (units of mmol/m^3)
- OX_200_ENSEMBLE_SD: Standard deviation (ensemble spread) of OX_200_ENSEMBLE
- OX_300_ENSEMBLE: Ensemble mean of dissolved oxygen at 300 m depth (units of mmol/m^3)
- OX_300_ENSEMBLE_SD: Standard deviation (ensemble spread) of OX_300_ENSEMBLE
- OX_300_GFDL: Spatial mean of hypoxic depth (2000-2100) from ROMS_GFDL
- OX_300_IPSL: Spatial mean of hypoxic depth (2000-2100) from ROMS_IPSL
- OX_300_HAD: Spatial mean of hypoxic depth (2000-2100) from ROMS_HAD
- OX_400_ENSEMBLE: Ensemble mean of dissolved oxygen at 400 m depth (units of mmol/m^3)
- OX_400_ENSEMBLE_SD: Standard deviation (ensemble spread) of OX_400_ENSEMBLE
- OX_500_ENSEMBLE: Ensemble mean of dissolved oxygen at 500 m depth (units of mmol/m^3)
- OX_500_ENSEMBLE_SD: Standard deviation (ensemble spread) of OX_500_ENSEMBLE
File: temp_iso.csv
Description: Downscaled climate projections (2000-2100) of temperature metrics averaged over our study domain (see Figure 1 in manuscript). Ensemble mean and spread were calculated as the mean and standard deviation (sd) of the 3 projections from ROMS-GFDL, ROMS-IPSL and ROMS-HADL (see manuscript for full descriptions).
Variables
- TEMP_150_ENSEMBLE: Ensemble mean of temperature at 150 m depth (degrees C)
- TEMP_150_ENSEMBLE_SD: Standard deviation (ensemble spread) of TEMP_150_ENSEMBLE
- TEMP_150_GFDL: Spatial mean of temperature at 150 m (2000-2100) from ROMS_GFDL
- TEMP_150_IPSL: Spatial mean of temperature at 150 m (2000-2100) from ROMS_IPSL
- TEMP_150_HAD: Spatial mean of temperature at 150 m (2000-2100) from ROMS_HAD
- DEPTH_ISO_26_5_ENSEMBLE: Ensemble mean of the depth of the 26.5 isopycnal (units of m)
- DEPTH_ISO_26_5_ENSEMBLE_SD: Standard deviation (ensemble spread) of DEPTH_ISO_26_5_ENSEMBLE
- DEPTH_ISO_26_5_GFDL: Spatial mean of isopycnal depth (2000-2100) from ROMS_GFDL
- DEPTH_ISO_26_5_IPSL: Spatial mean of isopycnal depth (2000-2100) from ROMS_IPSL
- DEPTH_ISO_26_5_HAD: Spatial mean of isopycnal depth (2000-2100) from ROMS_HAD
File: meso_hab.csv
Description: Downscaled climate projections (2000-2100) of mesopelagic habitat boundaries averaged over our study domain (see Figure 1 in manuscript). Ensemble mean and spread were calculated as the mean and standard deviation (sd) of the 3 projections from ROMS-GFDL, ROMS-IPSL and ROMS-HADL (see manuscript for full descriptions).
Variables
- MESO_HABITAT_ENSEMBLE: Ensemble mean of the distance (in units of m) from the upper mesopelagic boundary (depth at which light = 0.0217 W/m2) above and hypoxic boundary depth (depth at which DO = 63 mmol/m3) below
- MESO_HABITAT_ENSEMBLE_SD: Standard deviation (ensemble spread) of MESO_HABITAT_ENSEMBLE
- MESO_HABITAT_GFDL: Spatial mean of vertical mesopelagic habitat extent (2000-2100) from ROMS_GFDL
- MESO_HABITAT_IPSL: Spatial mean of mesopelagic habitat extent (2000-2100) from ROMS_IPSL
- MESO_HABITAT_HAD: Spatial mean of mesopelagic habitat extent (2000-2100) from ROMS_HAD
File: mesohabitat_proj_2000-2100.nc
Description: High resolution downscaled climate projections. We used a 1/30° model domain nested within a set of existing downscaled regional climate projections at 1/10° resolution for the broader California Current region. Briefly, the regional projections downscale three earth system model solutions (GFDL-ESM2M, IPSLCM5, and Hadley-GEM2-E) representing the physical and biogeochemical spread of the CMIP5 ensemble under the RCP8.5 high emissions scenario. The three corresponding high-resolution downscaled projections are hereafter referred to as ROMS-GFDL, ROMS-IPSL, and ROMS-HADL and depict low (GFDL), moderate (IPSL), and high (HADL) rates of warming under RCP8.5. This .nc file contains variables used to characterize changes in the vertical extent of mesopelagic habitat for the central California Current: the upper light boundary depth, the lower hypoxic boundary depth and dissolved oxygen concentrations at 300 m depth over our spatial domain in the central California Current defined as the latitudes from 34.5 to 40, bottom depths > 400 m and out to 200 km from the coastline.
Variables
- depth_isolume_gfdl: the upper boundary of the mesopelagic zone, defined as the depth (in units of m) at which light intensity reaches 0.0217 W/m2. These data are high resolution downscaled climate projections (2000-2100) from ROMS-GFDL for our entire spatial domain (see Fig.1).
- depth_isolume_ipsl: the upper boundary of the mesopelagic zone, defined as the depth (in units of m) at which light intensity reaches 0.0217 W/m2. These data are high resolution downscaled climate projections (2000-2100) from ROMS-IPSL for our entire spatial domain (see Fig.1).
- depth_isolume_hadl: the upper boundary of the mesopelagic zone, defined as the depth (in units of m) at which light intensity reaches 0.0217 W/m2. These data are high resolution downscaled climate projections (2000-2100) from ROMS-HADL for our entire spatial domain (see Fig.1).
- depth_hypoxic_gfdl: the lower boundary of the mesopelagic zone, defined as the depth (in units of m) at which dissolved oxygen reaches 63 mmol/m3 (hypoxic boundary). These data are high resolution downscaled climate projections (2000-2100) from ROMS-GFDL for our entire spatial domain (see Fig.1).
- depth_hypoxic_ipsl: the lower boundary of the mesopelagic zone, defined as the depth (in units of m) at which dissolved oxygen reaches 63 mmol/m3 (hypoxic boundary). These data are high resolution downscaled climate projections (2000-2100) from ROMS-IPSL for our entire spatial domain (see Fig.1).
- depth_hypoxic_hadl: the lower boundary of the mesopelagic zone, defined as the depth (in units of m) at which dissolved oxygen reaches 63 mmol/m3 (hypoxic boundary). These data are high resolution downscaled climate projections (2000-2100) from ROMS-HADL for our entire spatial domain (see Fig.1).
- do_300m_gfdl: Concentration of dissolved oxygen (in units of mmol/m^3) at 300 m depth. These data are high resolution downscaled climate projections (2000-2100) from ROMS-GFDL for our entire spatial domain (see Fig.1).
- do_300m_ipsl: Concentration of dissolved oxygen (in units of mmol/m^3) at 300 m depth. These data are high resolution downscaled climate projections (2000-2100) from ROMS-IPSL for our entire spatial domain (see Fig.1).
- do_300m_hadl: Concentration of dissolved oxygen (in units of mmol/m^3) at 300 m depth. These data are high resolution downscaled climate projections (2000-2100) from ROMS-HADL for our entire spatial domain (see Fig.1).
Code/software
Background: The following code was written by II in the R language (R version 4.4.1 (2024-06-14) -- "Race for Your Life" Copyright (C) 2024 The R Foundation for Statistical Computing) using RStudio (Version 2024.09.0+375 (2024.09.0+375) Copyright (C) 2024 by Posit Software, PBC). To run this code, the five .csv files available from Dryad (light.csv, meso_hab.csv, n2_no3_chl.csv, oxygen.csv, temp_iso.csv) must be downloaded and stored in your current working directory under the folder name "roms." There is one RMarkdown file (Projections_analysis_R.Rmd) that provides code for analyzing model output.
Objectives of code files:
Projections_analysis_R.Rmd
- Read in model output for oceanographic metrics and mesopelagic habitat boundary metrics
- Create plots of the spatial mean of ensemble means and spread of oceanographic variables (temperature at 150 m, oxygen at 300 m, integrated chlorophyll in upper 100 m, max buoyancy frequency (in upper 100), light at 150 m, mean nitrate (NO3) in upper 100 m) from 2000-2100.
- Create plot of changes in vertical mesopelagic habitat from 2000-2100
- Calculate summary statistics to describe changes in ocean conditions and mesopelagic habitat boundaries from the beginning (2000-2030) and end (2070-2100) of the century.
Access information
Access information
Other publicly accessible locations of data used in this study: Our analysis also uses publicly available data available at: "A high-resolution synthesis dataset for multistressor analyses along the US West Coast" by Kennedy et al. 2024 (https://doi.org/10.5194/essd-16-219-2024).
Downscaled Regional Climate Projections
We use an ensemble of high-resolution regional climate projections representing three different rates of warming under the RCP 8.5 high emissions scenario for the period 2000-2100. The earth system model solutions are first downscaled to 1/10° (~10 km) resolution for the broader California Current (30-48 °N)1,2 and subsequently nested at 1/30° (~3 km) resolution for the central California Current (32-44 °N) to improve the representation of local scale coastal upwelling processes3. In this approach, the 1/10° downscaled projections are forced directly by the CMIP5 earth system models using a time-varying delta method, and the high-resolution nested projections are subsequently forced by the downscaled projections using an offline nesting method (Fig. 1). The three earth system model solutions selected here are GFDL-ESM2M, IPSLCM5, and Hadley-GEM2-E as they include marine biogeochemical fields and represent the spread of the CMIP5 ensemble (GFDL-ESM2M = low rate of warming; IPSLCM5 = rate of warming close to ensemble mean; Hadley-GEM2-E = high rate of warming). While the RCP8.5 scenario may arguably depict an “unrealistically” warm future, we consider it useful here for two reasons: (1) it allowed exploring changes in mesopelagic habitat properties under extreme warming (e.g., identify potential thresholds), and (2) the lowest rate of warming (GFDL-ESM2M) is representative of the high end of the more moderate RCP4.5 scenario.
The physical and biogeochemical fields for both downscaled (1/10°) and nested (1/30°) projections are generated using an implementation of the Regional Ocean Modeling System (ROMS)4,5 for the California Current System coupled to NEMUCSC, a customized version of the North Pacific Ecosystem Model for Understanding Regional Oceanography (NEMURO)6. NEMUCSC includes three limiting macronutrients (nitrate, ammonium, and silicic acid), two phytoplankton functional groups (nanophytoplankton and diatoms), three zooplankton size-classes (microzooplankton, copepods, and euphausiids), three detritus pools (dissolved and particulate organic nitrogen and particulate silica), as well as carbon and oxygen cycling7,8. The three high-resolution nested projections of the coupled ROMS-NEMUCSC model are hereafter referred to as “ROMS-GFDL”, “ROMS-IPSL”, and “ROMS-HADL”.
References
1. Pozo Buil, M. et al. A Dynamically Downscaled Ensemble of Future Projections for the California Current System. Front. Mar. Sci. 8, 612874 (2021).
2. Fiechter, J., Pozo Buil, M., Jacox, M. G., Alexander, M. A. & Rose, K. A. Projected Shifts in 21st Century Sardine Distribution and Catch in the California Current. Front. Mar. Sci. 8, 685241 (2021).
3. Fiechter, J., Edwards, C. A. & Moore, A. M. Wind, Circulation, and Topographic Effects on Alongshore Phytoplankton Variability in the California Current. Geophys. Res. Lett. 45, 3238–3245 (2018).
4. Haidvogel, D. B. et al. Ocean forecasting in terrain-following coordinates: Formulation and skill assessment of the Regional Ocean Modeling System. J. Comput. Phys. 227, 3595–3624 (2008).
5. Shchepetkin, A. F. & McWilliams, J. C. The regional oceanic modeling system (ROMS): a split-explicit, free-surface, topography-following-coordinate oceanic model. Ocean Model. 9, 347–404 (2005).
6. Kishi, M. J. et al. NEMURO—a lower trophic level model for the North Pacific marine ecosystem. Ecol. Model. 202, 12–25 (2007).
7. Fiechter, J., Santora, J. A., Chavez, F., Northcott, D. & Messié, M. Krill Hotspot Formation and Phenology in the California Current Ecosystem. Geophys. Res. Lett. 47, e2020GL088039 (2020).
8. Cheresh, J. & Fiechter, J. Physical and Biogeochemical Drivers of Alongshore pH and Oxygen Variability in the California Current System. Geophys. Res. Lett. 47, e2020GL089553 (2020).