-------------------
GENERAL INFORMATION
-------------------


1. Title of Dataset:  
Data/Software supplement for “Processes that influence bottom temperatures in the California Current System” JGR–Oceans (in press) 2025


2. Authors: Michael Alexander, James D Scott, Mike Jacox, Dillon Amaya, Leah Wilczynski


3. Contact information: james.d.scott@noaa.gov


4. Date of data collection:
September, 2024


--------------------------
SHARING/ACCESS INFORMATION
-------------------------- 


1. Licenses/restrictions placed on the data:
CCO


2. Links to publications that cite or use the data:
Data/Software supplement for “Processes that influence bottom temperatures in the California Current System” JGR–Oceans (in press) 2025
Michael Alexander, James D Scott, Mike Jacox, Dillon Amaya, Leah Wilczynski



3. Recommended citation for the data:
Michael Alexander, James D Scott, Mike Jacox, Dillon Amaya, Leah Wilczynski. Data/Software supplement for “Processes that influence bottom temperatures in the California Current System” JGR–Oceans (in press) 2025. https://doi.org/10.25810/0j9m-4y38 



---------------------
DATA & FILE OVERVIEW
---------------------
The dataset size is approximately 17.5 GB. the netCDF files are available to download through Globus. See "Additional Information" below.

Software-specific information:
 
.NCL files use the NCAR command language:
The NCAR Command Language (Version 6.6.2) [Software]. (2024).
Boulder, Colorado: NCAR/CISL/VETS. http://dx.doi.org/10.5065/D6WD3XH5
 
.nc (netCDF) files can also be accessed with the following software:
ncview : https://anaconda.org/conda-forge/ncview
Python : https://www.python.org/
MatLab: https://www.mathworks.com/products/matlab.html
 
The codes to compute mixed layer depth use the temperature and salinity from the GLORYS12v1 reanalysis data (Lellouche et al., 2021) which are freely available at: https://data.marine.copernicus.eu/product/GLOBAL_MULTIYEAR_PHY_001_030/services



1. File List:

A. Alexander_Data-NCL-files.zip (86.85 KB) contains NCAR command language code files:

Fig.1.ncl
Fig.2.ab.ncl
Fig.3.ncl
Fig.4.ncl
Fig.5.ncl
Fig.6.ncl
Fig.7.ncl
Fig.8.ncl
Fig.9.ncl
Fig.10.ncl
Fig.11.ncl
Fig.12.ncl
Fig.13.ncl
Fig.14.ncl
Fig.S1.ncl
Fig.S2.S3.ncl
Fig.S4.S5.ncl
Fig.S6.ncl
Fig.S7.ncl
Fig.S11.ncl
Fig.S12.ncl
Fig.S13.ncl
Fig.S14.ncl
Fig.S15.ncl
mld.from.daily.rho.ncl
thermocline.depth.VRI.ncl


B. Alexander_Data-NC_NetCDF_files.zip (17.43 GB) contains netCDF data files used in the above code to generate the analyses in the JGR-Oceans journal article above:

Note: This file is available to download through Globus. See "Additional Information" below.

AIC.scores1.new.nc
AIC.scores2.new.nc
AIC.scores3.new.nc
AIC.scores4.new.nc
AIC.scores5.new.nc
AIC.scores6.new.nc
botT.daily.mean.1993-2019.CalCS.nc
botT.daily.mean.clim.1993-2019.CalCS.nc
bottomT.mon.mean.199301-201912.nc
botU.mon.mean.1993-2019.CalCS.nc
botV.mon.mean.1993-2019.CalCS.nc
bwt.acc.pvals.nc
bwt.thermdepth.botV.seas.MLR.coast.nc
bwt.thermdepth.botV.seas.MLR.coast.region.avg.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.0-100m.apct.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.0-100m.avg.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.100-400m.apct.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.100-400m.avg.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.apct.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.avg.nc
ccs.coastal.longitudes.nc
dtdz.mon.mean.1993-2019.10m.nc
fig_data_bta_time_series.nc
fig_data_ek_ssh_correlations.nc
GLO-MFC_001_030_mask_bathy.nc*
GLORYS_CUTI_24-48N.nc
LME.GLORYS.mask.nc
mld.mon.mean.1993-2019.nc
nino.to.vo.regr.nc
sst.daily.mean.1993-2019.CalCS.nc
sst.daily.mean.clim.1993-2019.CalCS.nc
sst.mon.mean.199301-201912.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
thetao.mon.mean.1993-2019.CalCS.nc
zos.daily.mean.1993-2019.CalCS.nc
zos.daily.mean.clim.1993-2019.CalCS.nc
zos.mon.mean.199301-201912.nc


Additional Information

To download this data set (Alexander_Data-NC_NetCDF_files.zip) with Globus, click on the link in the Related URL field on CU Scholar or go to https://app.globus.org/file-manager?origin_id=28a91ff0-c071-433e-8d57-e1119215807d&origin_path=%2F, and follow these instructions.

	Step 1: Sign in to Globus (https://www.globus.org/app/login). If you are with an institution that is registered with Globus, you can simply sign in with your institutional credentials. If your institution is not registered with Globus, you will need to first make an account (https://www.globusid.org/create) before signing in.
	
	Step 2: Install Globus Connect Personal (https://www.globus.org/globus-connect-personal) on your local computer. This will establish your local endpoint to which you can download the CU Scholar dataset. You'll give the local endpoint a name (e.g., "Jane Doe's Laptop")
	
	Step 3: Click on the Globus URL for the dataset on its CU Scholar landing page. This will bring up the dataset on the left hand side of the Globus file manager. On the right hand side, search for your endpoint name (e.g., "Jane Doe's Laptop"), which will bring up the file system on your local machine (make sure Globus Connect Personal is running). Highlight the dataset files that you want to transfer and click on the blue "start" button to transfer the data to your machine.
	
	Data storage for this data set is supported by CU Boulder Research Computing's PetaLibrary.



2. Relationship between files:       


A. NCL code to reproduce main figures in the Article:


Fig.1.ncl requires netCDF data files:
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
 Figure. 1. Depth of the ocean’s bottom in the CCS domain. GLORYS bathymetry (m) on the shelf off the west coast of Baja California and the contiguous US, approximately 23°-50°N. (a) The bottom depth on the shelf (< 400 m) on the model grid as a function of latitude and distance from the coastline and (b) in a traditional map view for the north (N), north central (NC), south central (SC), and south (S) regions. Half of the latitudes have a shelf width between 21-58 km.


Fig.2.ab.ncl requires netCDF data files:
LME.GLORYS.mask.nc
GLO-MFC_001_030_mask_bathy.nc
Caption:
Figure 2. Time series of SST and BWT anomalies in the four CCS regions and the correlations between regions for each variable separately. Area average of the three month running mean of the (a) SST and (b) BWT anomalies (℃) on the continental shelf (< 400 m deep) for the N (green), NC (black), SC (red) and south (blue) regions. The correlation between monthly SST and BWT anomalies 274 in each of the four regions is given in (a). (c) Anomaly correlations between regions for SST (red, upper left) and BWT (blue lower right).


Fig.3.ncl requires netCDF data files:
ccs.coastal.longitudes.nc
bottomT.mon.mean.199301-201912.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
mld.mon.mean.1993-2019.nc
dtdz.mon.mean.1993-2019.10m.nc
botV.mon.mean.1993-2019.CalCS.nc
GLO-MFC_001_030_mask_bathy.nc
Caption: 
Figure 3. Interannual standard deviation (𝝈) of BWT (°C) on the continental shelf for (a) JFM and (b) JAS.


Fig.4.ncl requires netCDF data files:
ccs.coastal.longitudes.nc
bottomT.mon.mean.199301-201912.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
mld.mon.mean.1993-2019.nc
dtdz.mon.mean.1993-2019.10m.nc
botV.mon.mean.1993-2019.CalCS.nc
botU.mon.mean.1993-2019.CalCS.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Figure 4. Correlations between BWT anomalies and six ocean variables at each model grid square on the shelf. The seasonal average anomaly correlation coefficient (ACC) for BWT with (a,g) Mixed Layer Depth (MLD), (b,h) thermocline depth (TD), (c,i) thermocline gradient (∇TThrm), (d, j) vertical bottom temperature gradient (∇TBot), (e, k) bottom meridional velocity (VBot), and (f, l) and bottom zonal velocity (UBot) for (top) JFM and (bottom) JAS. Black lines delineate the four regions defined in Figure1; black squares indicate where the thermocline is undefined and gray squares where the correlations are not significant based on the t-statistic at the 95% confidence level.


Fig.5.ncl requires netCDF data files:
ccs.coastal.longitudes.nc
bottomT.mon.mean.199301-201912.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
mld.mon.mean.1993-2019.nc
dtdz.mon.mean.1993-2019.10m.nc
botV.mon.mean.1993-2019.CalCS.nc
botU.mon.mean.1993-2019.CalCS.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Figure 5. Values from a multilinear regression model from six predictor ocean variables BWT anomalies. Standardized regression coefficients from a Multilinear Regression (MLR) model for the predictors: (a, g) MLD, (b, h) TD, (c, i) ∇TBot, (d ,j) ∇TBot, (e, k) VBot, and (f, l) UBot for (top) JFM and (bottom) JAS. Colored areas indicate significant regression coefficients (90%) and gray areas are not significant. Black squares indicate where the thermocline is undefined; black lines delineate the four regions defined in Figure 1.


Fig.6.ncl requires netCDF data files:
bwt.thermdepth.botV.seas.MLR.coast.region.sig.apct.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.0-100m.apct.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.100-400m.apct.nc
Caption:
Figure 6. Percentage of the shelf area over the four regions with significant regression
coefficients shown in Figure 5 for the (a, d) entire shelf (≤ 400 m deep), (b, e) near shore (< 100m deep), and (c, f) offshore (100-400 m deep) in (top) JFM and (bottom) JAS.


Fig.7.ncl requires netCDF data files:
bwt.thermdepth.botV.seas.MLR.coast.region.sig.avg.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.0-100m.avg.nc
bwt.thermdepth.botV.seas.MLR.coast.region.sig.100-400m.avg.nc
Caption:
Figure 7. Area average over the four regions of the significant regression coefficients shown in Figure 5 for the (a, d) entire shelf, (b,e) near shore, and (c, f) offshore during (top) JFM and (bottom) JAS.


Fig.8.ncl requires netCDF data files:
botT.daily.mean.1993-2019.CalCS.nc
botT.daily.mean.clim.1993-2019.CalCS.nc
zos.daily.mean.1993-2019.CalCS.nc
zos.daily.mean.clim.1993-2019.CalCS.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Figure 8. The relationship between SSH at the southern end of the domain and SSH and BWT anomalies within the domain as a function of frequency/period, which is useful for identifying propagating features, including coastally trapped waves (CTWs). Cross spectra showing the (a,d) Coherence, (b, e) phase and (c, f) lag in days between the southern Baja California (“Cabo”, 23°N-23.5°N) SSH and coastal (nearest 0.5° longitude to shore) for (a,b,c) SSH and (d, e, f) BWT (< 400m) as a function of latitude. A positive phase or lag indicates that the Cabo SSH index leads the SSH and BWT at higher latitudes. Gray indicates points where the coherence is not significant (at the 90% level).


Fig.9.ncl requires netCDF data files:
botT.daily.mean.1993-2019.CalCS.nc
botT.daily.mean.clim.1993-2019.CalCS.nc
bottomT.mon.mean.199301-201912.nc
zos.daily.mean.1993-2019.CalCS.nc
zos.daily.mean.clim.1993-2019.CalCS.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Figure 9. Lead-lag correlations as a function of latitude for time filtered (20-180 day band pass) coastal (nearest 0.5° longitude to shore) averaged anomalies between Cabo SSH anomalies and (a) SSH anomalies and (b) BWT (≤ 400m) anomalies and (c) between the collocated SSH and BWT anomalies.


Fig.10.ncl requires netCDF data files:
zos.mon.mean.199301-201912.nc
bottomT.mon.mean.199301-201912.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Figure 10. Evolving basin-wide SSH anomaly patterns that are related to BWTs at the southern end of the CCS reveal ENSO’s influence. Correlations between the Cabo BWT anomalies averaged across the shelf (23°N-23.5°N, 0.5° longitude nearest the coast, ≤ 400m deep) during NDJ with the monthly mean SSH anomalies over the Pacific from 10°S to 60°N at lags of (a) -9, (b) -6, (c) -3, (d) 0, (e) 3, and (f) 6 months, which range from the previous March to the following June. A negative lag indicates the SSH anomalies precede the Cabo BWT index. The location of the Cabo index is depicted by a triangle in panel (a).


Fig.11.ncl requires netCDF data files:
ccs.coastal.longitudes.nc
sst.mon.mean.199301-201912.nc
bottomT.mon.mean.199301-201912.nc
GLO-MFC_001_030_mask_bathy.nc
Caption:
Figure 11. The influence of ENSO on BWT anomalies on the shelf in the CCS. Monthly BWT anomalies on the shelf between 23°N-50°N regressed on the NDJ Niño 3.4 (5°S-5°N, 170°W-120°W) SST anomalies. Regression values (℃ / Nino SST 𝞼) are for shown for (a) June prior to the Nino index (-6 month lag) to the following (j) March (+3 month lag).


Fig.12.ncl requires netCDF data files:
GLORYS_CUTI.ekmanonly.nc
GLORYS_CUTI_24-48N.nc
botT.daily.mean.1993-2019.CalCS.nc
botT.daily.mean.clim.1993-2019.CalCS.nc
zos.daily.mean.clim.1993-2019.CalCS.nc
zos.daily.mean.1993-2019.CalCS.nc
sst.daily.mean.clim.1993-2019.CalCS.nc
sst.daily.mean.1993-2019.CalCS.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Fig_data_ek_ssh_correlations.nc
Caption:
Figure 12. The influence of local winds and poleward propagating CTWs on SSH, BWT and SST anomalies in the CCS. ACCs as a function of latitude averaged over the continental shelf (< 400m) for (a) SSH, (b) BWT, (c) SST with the SSH anomalies at 30.5°N-31°N (pink) and the local wind-driven component of the Coastal Upwelling Transport Index (-1*CUTI, blue) for GLORYS (solid) and ROMS (dashed). The ACC values are computed daily based on 8-day averages (the archived frequency of the ROMS output), where the anomalies lag the SSH and CUTI indices by one 8-day period. Given the large number of samples,results in low ACC values (<0.1) being statistically significant based on two sided t-test at a 95% confidence level, even after taking the
autocorrelation in the variables into account.


Fig.13.ncl requires netCDF data files:
Fig_data_bta_time_series.nc
Caption:
Figure 13. The influence of wind forcing and ocean boundary conditions, including CTWs, on BWTs in the CCS obtained using model sensitivity studies. Time series of BWT anomalies on the shelf (< 400 m) averaged the North, North Central, and South Central regions. The Control with full forcing, and variable Wind and Ocean sensitivity simulations (in which the other forcing fields are set to their respective climatology) are shown by gray, blue and pink lines, respectively. The correlation between the control and the sensitivity run is shown in the upper right corner of each panel.


Fig.14.ncl requires netCDF data files:
nino.to.vo.regr.nc
thetao.mon.mean.1993-2019.CalCS.nc
mld.mon.mean.1993-2019.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
GLO-MFC_001_030_mask_bathy.nc
dtdz.mon.mean.1993-2019.10m.nc
bottomT.mon.mean.199301-201912.nc
Caption:
Figure 14. The climatology and changes in the ocean temperature as a function of depth west of Baja California illustrating the large changes in MLD, TD and vertical temperature gradients during a strong El Niño event. Cross sections at 26°N during JFM of the temperature (°C, shading), MLD (dashed line) and TD (solid line) for (a) climatology and (b) 2016. (c) Temperature anomalies during JFM 2016 (°C, shading) and the MLD (dashed) and TD (solid) for the climatology (black) and during 2016 (gray).


B. NCL code to produce supplemental figures:


Fig.S1.ncl requires netCDF data files:
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Fig. S1. GLORYS seasonal mean MLD computed as the depth with a 0.125 (kg/m3) density increase from the surface value in top) JFM and bottom) JAS.


Fig.S2.S3.ncl requires netCDF data files:
nino.to.vo.regr.nc
thetao.mon.mean.1993-2019.CalCS.nc
mld.mon.mean.1993-2019.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
GLO-MFC_001_030_mask_bathy.nc
Caption:
Fig. S2. Thermocline depth variability: full range (light shading) and interquartile range (dark shading) and the temperature climatology (contours) in JFM (°C) for zonal cross sections at: a) 46.5°N, b) 44°N, c) 38°N, d) 34.5°N, e) 28.5°N, and f) 26°N.
Fig. S3. Same as Fig. S2 but for JAS.


Fig.S4.S5.ncl requires netCDF data files:
nino.to.vo.regr.nc
thetao.mon.mean.1993-2019.CalCS.nc
mld.mon.mean.1993-2019.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
GLO-MFC_001_030_mask_bathy.nc
Caption:
Fig. S4. Thermocline depth variability: full range (light shading) and interquartile range (dark shading) and the interannual temperature standard deviation (𝝈) (contours) in JFM (°C) for zonal cross sections at: a) 46.5°N, b) 44°N, c) 38°N, d) 34.5°N, e) 28.5°N, and f) 26°N.
Fig. S5. Same as Fig. S4 but for JAS.


Fig.S6.ncl requires netCDF data files:
ccs.coastal.longitudes.nc
bottomT.mon.mean.199301-201912.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Fig S6. Anomaly correlation coeWicient (ACC) between the SST and BWT anomalies averaged over a) JFM and b) JAS. Values are shown where the bottom depth is ≤ 400 m. Gray indicates that the grid square correlation values that were not significant at the 95% confidence level based on a t-test.


Fig.S7.ncl requires netCDF data files:
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
botV.mon.mean.1993-2019.CalCS.nc
Caption:
Fig. S7 Meridional Velocity (cm s-1) at the lowest reanalysis level on the continental shelf (< 400m depth) for (a) JFM, (b) JAS.


Fig.S11.ncl requires netCDF data files:
ccs.coastal.longitudes.nc
bottomT.mon.mean.199301-201912.nc
thermocline.VRI.mon.mean.1993-2019.CalCS.new.nc
mld.mon.mean.1993-2019.nc
dtdz.mon.mean.1993-2019.10m.nc
botV.mon.mean.1993-2019.CalCS.nc
botU.mon.mean.1993-2019.CalCS.nc
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Fig. S11. Correlation between the BWT anomalies predicted by the MLR (shown in Fig.5) and the actual BWT anomalies for a) JFM , b) JAS. Correlations that are not statistically significant (at the 95% confidence) are gray.


Fig.S12.ncl requires netCDF data files:
AIC.scores1.new.nc
AIC.scores2.new.nc
AIC.scores3.new.nc
AIC.scores4.new.nc
AIC.scores5.new.nc
AIC.scores6.new.nc
Caption:
Fig. S12. Probability distribution of AIC scores for all bottom temperature locations in Fig. 5. The different colors represent the distributions for including 1(black), 2(blue), 3(green), 4(red), 5(purple) 6 (orange) predictor variables in the multiple linear regressions for BWT anomalies. Median and average scores drop with higher order MLRs; a MLR with a lower AIC scores indicate that it better fits the data.


Fig.S13.ncl requires netCDF data files:
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Fig. S13. Correlations between the southern Baja California (“Cabo”) SSH anomalies (23°N-23.5°N) with monthly mean SSH anomalies at lead/lags of a) -9, b) -6, c) -3, d) 0, e)3, and f) 6 months (f). A negative lag indicates that the basin SSH anomalies lead those at Cabo (triangle in (a)).


Fig.S14.ncl requires netCDF data files:
GLO-MFC_001_030_mask_bathy.nc
LME.GLORYS.mask.nc
Caption:
Fig. S14. Same as Fig. S12 but for correlations of the Cabo SSH index with monthly mean SST anomalies.


Fig.S15.ncl requires netCDF data files:
GLORYS_CUTI.ekmanonly.nc
GLORYS_CUTI_24-48N.nc
botT.daily.mean.1993-2019.CalCS.nc
botT.daily.mean.clim.1993-2019.CalCS.nc
Caption:
Fig. S15 ACCs between the Cabo SSH index at 8 day lag constructed from anomalies along the continental shelf (< 400m) for (a) BWT, (b) SST, (c) SST using SSH index (30.5°N-31°N, pink) and using local -1*CUTI index at each latitude (blue) from GLORYS.


C. NCL code to compute mixed layer depth from density:
mld.from.daily.rho.ncl


D. NCL code to compute thermocline depth from temperature:
thermocline.depth.VRI.ncl