Variable preparation

Introduction

This vignette walks through the steps needed to create mean synoptic states during all of the MHWs detected in the previous vignette. These synoptic states consist of the variables in the ocean heat budget equation.

After first going through this entire process with the NAPA model output it became clear that this analysis should actually be run with obs/reanalysis data, and the NAPA model results may be compared against this as desired. To this end we will use the following variables as found in the following data products:

• NOAA: (optimally interpolated) SST
• GLORYS: Mixed layer depth (metres), U & V current vectors (~0.5m)
• ERA 5: U & V wind vectors (10m), air temperature (2m), net heat flux:
  • latent & sensible heat flux, shortwave & longwave radiation

All of these products and variables have a shared daily resolution period from 1993 – 2018. The spatial resolutions for NOAA and ERA 5 are 1/4 degrees and the GLORYS data are 1/12 degree. The coordinates are slightly off between OISST and ERA 5 & GLORYS, with the coordinate system for NOAA being centred per pixel while the other two products have their coordinate system based on the northwest corner. In order to compare the OISST dataset to the other 1/4 degree products the lat values for OISST will have 0.125 added to them, and the longitude values will have 0.125 subtracted from them. The ERA5 data are hourly, so will need to be binned into mean daily values. In order to trade between efficiency and accuracy we will be converting everything to the 1/4 degree grid of the GLORYS/ERA 5 data at a daily temporal resolution.

# Packages used in this vignette
library(jsonlite, lib.loc = "../R-packages/") # Necessary for some version mismatches...
library(tidyverse) # Base suite of functions
library(tidync, lib.loc = "../R-packages/")
library(lubridate) # For convenient date manipulation

# Set number of cores
doMC::registerDoMC(cores = 50)

# Disable scientific notation for numeric values
  # I just find it annoying
options(scipen = 999)

# Corners of the study area
NWA_corners <- readRDS("data/NWA_corners.Rda")

# The region pixel lon/lat values
NWA_info <- readRDS("data/NWA_info.Rda")

# Load MHW results
OISST_region_MHW <- readRDS("data/OISST_region_MHW.Rda")

# MHW Events
MHW_event <- OISST_region_MHW %>%
  select(-cats) %>%
  unnest(events) %>%
  filter(row_number() %% 2 == 0) %>%
  unnest(events)

Base datasets

Rather than go about performing all of the calculations below on piecemeal bits of data, we will load each variable into memory at once. The following code chunk shows how we create functions that load each of the variables from the different data products. We will then create and save this complete dataset for ease of use later on.

# Load the functions used below
  # I've chosen to house the functions  in the following script
  # in order to keep this vignette tidier
source("code/functions.R")
# See the code/workflow.R script for the exact code used for this step

# The code used to perform these steps may be seen in code/workflow.R

Climatologies

In the data packets we need to create for the SOM we will need to include anomaly values. In order to do this we need to first create daily climatologies for each variable for each pixel. To do so we will need to load all of the files and then pixel-wise go about getting the seasonal (mean daily) climatologies. This will be done with the same function (ts2clm()) that is used for the MHW climatologies. One-by-one we will load an entire dataset (created above) into memory so that we can perform the necessary calculations. Hold on to your hats, this is going to be RAM heavy…

# The code used to create the climatologies may be found in code/workflow.R

Anomalies

The last step before we can begin creating our data packets is to subtract our climatology data from our base data for each of the variables used in this study. We then save these large anomaly cubes to allow for easier synoptic state creation later on.

# The code used to create the rest of the anomalies may be found in code/workflow.R

Synoptic states

The next step is to create mean synoptic states for each variable during each of the MHWs detected in each region. To do this we simply load the combined anomaly dataset of all of our chosen variables and slice off only the bits we need during each of the observed MHWs. These slices are then meaned, pixel-wise, to create the synoptic states during each event.

# This code may be found in code/workflow.R

With all of our mean synoptic states created it is now time to feed them to the Self-organising map (SOM).

Session information

sessionInfo()

R version 3.6.1 (2019-07-05)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 16.04.6 LTS

Matrix products: default
BLAS:   /usr/lib/openblas-base/libblas.so.3
LAPACK: /usr/lib/libopenblasp-r0.2.18.so

locale:
 [1] LC_CTYPE=en_CA.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_CA.UTF-8        LC_COLLATE=en_CA.UTF-8    
 [5] LC_MONETARY=en_CA.UTF-8    LC_MESSAGES=en_CA.UTF-8   
 [7] LC_PAPER=en_CA.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_CA.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] bindrcpp_0.2.2  lubridate_1.7.4 tidync_0.2.1    forcats_0.3.0  
 [5] stringr_1.3.1   dplyr_0.7.6     purrr_0.2.5     readr_1.1.1    
 [9] tidyr_0.8.1     tibble_1.4.2    ggplot2_3.0.0   tidyverse_1.2.1
[13] jsonlite_1.6   

loaded via a namespace (and not attached):
 [1] tidyselect_0.2.4  haven_1.1.2       lattice_0.20-35  
 [4] colorspace_1.3-2  htmltools_0.3.6   yaml_2.2.0       
 [7] rlang_0.2.2       R.oo_1.22.0       pillar_1.3.0     
[10] glue_1.3.0        withr_2.1.2       R.utils_2.7.0    
[13] RNetCDF_1.9-1     doMC_1.3.5        modelr_0.1.2     
[16] readxl_1.1.0      foreach_1.4.4     bindr_0.1.1      
[19] plyr_1.8.4        munsell_0.5.0     gtable_0.2.0     
[22] workflowr_1.1.1   cellranger_1.1.0  rvest_0.3.2      
[25] R.methodsS3_1.7.1 codetools_0.2-15  evaluate_0.11    
[28] knitr_1.20        parallel_3.6.1    broom_0.5.0      
[31] Rcpp_0.12.18      backports_1.1.2   scales_1.0.0     
[34] hms_0.4.2         digest_0.6.16     stringi_1.2.4    
[37] ncdf4_1.16.1      grid_3.6.1        rprojroot_1.3-2  
[40] cli_1.0.0         tools_3.6.1       magrittr_1.5     
[43] lazyeval_0.2.1    crayon_1.3.4      whisker_0.3-2    
[46] pkgconfig_2.0.2   xml2_1.2.0        iterators_1.0.10 
[49] assertthat_0.2.0  rmarkdown_1.10    httr_1.3.1       
[52] rstudioapi_0.7    R6_2.2.2          ncmeta_0.0.4     
[55] nlme_3.1-137      git2r_0.23.0      compiler_3.6.1