Project EDDIE: Modeling Climate Change Effects on Lakes Using Distributed Computing

Jemma Stachelek

2022-06-22

Project EDDIE is a series of activities entitled and developed by:

Modeling Climate Change Effects on Lakes Using Distributed Computing Module This module was initially developed by Carey, C.C., S. Aditya, K. Subratie, and R. Figueiredo. 1 May 2016. http://cemast.illinoisstate.edu/data-for-students/modules/lake-modeling.shtml (Module development was supported by NSF DEB 1245707 and ACI 1234983).

The following tutorial loosely follows the Project EDDIE steps with the exception of using the built-in run_example_sim function in the glmtools package to complete the initial set-up and file handling. This streamlines setup and avoids manual interaction with system files.

ACTIVITY A

OBJECTIVE 1: Download the GLM files and R packages successfully onto your computer.

Install and load packages

install.packages('sp')
install.packages('glmtools', repos=c('http://cran.rstudio.com', 'http://owi.usgs.gov/R'))

library(glmtools)

Verify setup

glm_version()

        ------------------------------------------------
       |  General Lake Model (GLM)   Version 2.2.0rc5    |
       ------------------------------------------------
--help  : show this blurb
--nml <nmlfile> : get parameters from nmlfile
--xdisp : display temp/salt and selected others in x-window
--xdisp <plotsfile> : like --xdisp, but use <plotsfile> instead of plots.nml
--saveall : save plots to png files
--save-all-in-one : save all plots to png file
[1] 0

OBJECTIVE 2: Migrate the example files that come with your downloaded GLM files onto a new directory on your computer

File handling

sim_folder <- "data"
dir.create(sim_folder, showWarnings = FALSE)  
run_example_sim(sim_folder)

## writing nml file to  data/glm3.nml

## Warning in glm.systemcall(sim_folder, glm_path, verbose, system.args): Custom
## path to GLM executable set via 'GLM_PATH' environment variable as: /home/jemma/
## Documents/Science/Models/GLM_docker/glm

## Warning in system2(glm_path, wait = TRUE, stdout = "", stderr = "", args =
## system.args): error in running command

## simulation complete. 
## *.nc output located in  data/output.nc

## [1] "data"

nml_file <- file.path(sim_folder, "glm2.nml")
nml <- read_nml(nml_file)

print(nml)

## &glm_setup
##    sim_name = 'GLMSimulation'
##    max_layers = 480
##    min_layer_vol = 0.025
##    min_layer_thick = 0.1
##    max_layer_thick = 3
##    Kw = 0.331
##    coef_mix_conv = 0.2
##    coef_wind_stir = 0.402
##    coef_mix_shear = 0.2
##    coef_mix_turb = 0.51
##    coef_mix_KH = 0.1
##    coef_mix_hyp = 0.2
## /
## &morphometry
##    lake_name = 'Sparkling'
##    latitude = 32
##    longitude = 35
##    bsn_len = 21000
##    bsn_wid = 13000
##    bsn_vals = 15
##    H = 301.712, 303.018285714286, 304.324571428571, 305.630857142857, 306.937142857143, 308.243428571429, 309.549714285714, 310.856, 312.162285714286, 313.468571428571, 314.774857142857, 316.081142857143, 317.387428571429, 318.693714285714, 320
##    A = 0, 45545.8263571429, 91091.6527142857, 136637.479071429, 182183.305428571, 227729.131785714, 273274.958142857, 318820.7845, 364366.610857143, 409912.437214286, 455458.263571429, 501004.089928571, 546549.916285714, 592095.742642857, 637641.569
## /
## &time
##    timefmt = 2
##    start = '2010-04-15 00:00:00'
##    stop = '2010-12-30 00:00:00'
##    dt = 3600
##    timezone = 7
## /
## &output
##    out_dir = '.'
##    out_fn = 'output'
##    nsave = 24
##    csv_lake_fname = 'lake'
##    csv_point_nlevs = 0
##    csv_point_fname = 'WQ_'
##    csv_point_at = 17
##    csv_point_nvars = 2
##    csv_point_vars = 'temp','salt','OXY_oxy'
##    csv_outlet_allinone = .false.
##    csv_outlet_fname = 'outlet_'
##    csv_outlet_nvars = 3
##    csv_outlet_vars = 'flow','temp','salt','OXY_oxy'
##    csv_ovrflw_fname = 'overflow'
## /
## &init_profiles
##    lake_depth = 18.288
##    num_depths = 3
##    the_depths = 0, 0.2, 18.288
##    the_temps = 3, 4, 4
##    the_sals = 0, 0, 0
##    num_wq_vars = 6
##    wq_names = 'OGM_don','OGM_pon','OGM_dop','OGM_pop','OGM_doc','OGM_poc'
##    wq_init_vals = 1.1, 1.2, 1.3, 1.2, 1.3, 2.1, 2.2, 2.3, 1.2, 1.3, 3.1, 3.2, 3.3, 1.2, 1.3, 4.1, 4.2, 4.3, 1.2, 1.3, 5.1, 5.2, 5.3, 1.2, 1.3, 6.1, 6.2, 6.3, 1.2, 1.3
## /
## &meteorology
##    met_sw = .true.
##    lw_type = 'LW_IN'
##    rain_sw = .false.
##    atm_stab = .false.
##    catchrain = .false.
##    rad_mode = 1
##    albedo_mode = 1
##    cloud_mode = 4
##    subdaily = .false.
##    meteo_fl = 'nldas_driver.csv'
##    wind_factor = 1
##    sw_factor = 1.08
##    lw_factor = 1
##    at_factor = 1
##    rh_factor = 1
##    rain_factor = 1
##    ce = 0.00132
##    ch = 0.0013
##    cd = 0.0013
##    rain_threshold = 0.01
##    runoff_coef = 0.3
## /
## &bird_model
##    AP = 973
##    Oz = 0.279
##    WatVap = 1.1
##    AOD500 = 0.033
##    AOD380 = 0.038
##    Albedo = 0.2
## /
## &inflow
##    num_inflows = 0
##    names_of_strms = 'Riv1','Riv2'
##    subm_flag = .false., .false.
##    strm_hf_angle = 65, 65
##    strmbd_slope = 2, 2
##    strmbd_drag = 0.016, 0.016
##    inflow_factor = 1, 1
##    inflow_fl = 'inflow_1.csv','inflow_2.csv'
##    inflow_varnum = 4
##    inflow_vars = 'FLOW','TEMP','SALT','OXY_oxy','SIL_rsi','NIT_amm','NIT_nit','PHS_frp','OGM_don','OGM_pon','OGM_dop','OGM_pop','OGM_doc','OGM_poc','PHY_green','PHY_crypto','PHY_diatom'
## /
## &outflow
##    num_outlet = 0
##    flt_off_sw = .false.
##    outl_elvs = -215.5
##    bsn_len_outl = 799
##    bsn_wid_outl = 399
##    outflow_fl = 'outflow.csv'
##    outflow_factor = 0.8
## /
## &snowice
##    snow_albedo_factor = 1
##    snow_rho_max = 300
##    snow_rho_min = 50
## /
## &sed_heat
##    sed_temp_mean = 4.5
##    sed_temp_amplitude = 0.25
##    sed_temp_peak_doy = 242.5
## /

Get Parameters of the nml control file

get_nml_value(nml, 'lake_name')

## [1] "Sparkling"

plot_meteo(nml_file)

OBJECTIVE 3: Run the model and look at output

run_glm(sim_folder, verbose=TRUE)

Examine model output

nc_file <- file.path("data/output.nc")
plot_temp(file = nc_file, fig_path = FALSE)

OBJECTIVE 4: Examine how your modeled GLM data compares to the observed field data for your lake.

nc_file <- file.path("data/output.nc")
field_path <- file.path(sim_folder, 'field_data.csv')
plot_var_compare(nc_file, field_path, 'temp', resample=FALSE)

## Warning: Removed 38 rows containing missing values (geom_raster).

List and compare available metrics

sim_metrics(with_nml = FALSE)

## [1] "buoyancy.freq"     "center.buoyancy"   "thermo.depth"     
## [4] "water.density"     "water.temperature"

compare_to_field(nc_file, field_path, metric="thermo.depth", 
                 as_value = TRUE, na.rm = TRUE)

##               DateTime       obs       mod
## 1  2010-04-27 01:00:00  7.992360  5.750947
## 2  2010-05-10 01:00:00 10.522474  9.806721
## 3  2010-05-24 01:00:00 10.703003  9.829644
## 4  2010-06-08 01:00:00 11.348012  9.775764
## 5  2010-06-21 01:00:00 11.585997  9.700781
## 6  2010-07-06 01:00:00  6.580984  9.475542
## 7  2010-07-20 01:00:00 10.261094  4.859414
## 8  2010-08-03 01:00:00  7.572742  6.323220
## 9  2010-08-17 01:00:00  7.691791  6.285066
## 10 2010-09-01 01:00:00 11.252144  7.595761
## 11 2010-09-14 01:00:00 10.429662  8.782602
## 12 2010-09-28 01:00:00 11.588422 11.059639
## 13 2010-10-12 01:00:00 12.782920 11.410977
## 14 2010-10-26 01:00:00 14.590728 13.204661

Plot modeled versus observed metric values

therm_depths <- compare_to_field(nc_file, field_path, 
                  metric="thermo.depth", as_value = TRUE, na.rm = TRUE)

plot(therm_depths$DateTime, therm_depths$obs, 
     type = "l", col = "blue", ylim = c(0, 32),
     ylab="Thermocline depth in meters")
lines(therm_depths$DateTime, therm_depths$mod, col = "red")
legend("topright",c("Observed", "Modeled"), lty = c(1, 1), 
     col = c("blue", "red"))

ACTIVITY B

OBJECTIVE 5: Develop a climate change scenario for your model lake

• Develop a climate scenario (it can be for any region!)

• Create a corresponding meteorological input file. Think through all of the components of the proposed scenario. For example, which of the meteorological variables (air temperature, precipitation, wind, etc.) will be modified and how? Will they be short-term or long-term modifications?

• Run the file and examine how it alters the physical structure of the lake. How does your climate scenario change the thermal structure of the lake? What does the temperature profile look like? How does the depth of the thermocline change? How does the timing of stratification and magnitude of evaporation change?

• Compare the modeled output to the observed. What are the implications of your climate scenario for future water quality and quantity?

• Create a few figures to highlight the results of your climate scenario. It would be helpful to present both the meteorological input file as well as the lake thermal plots so that you can see how the lake responded to your climate forcing.

Modify the gml2.nml file

• change model start and end to ‘2000-03-01 00:00:00’ and ‘2000-12-31 00:00:00’ respecitvely

ACTIVITY C

OBJECTIVE 6: Set up hundreds of GLM simulations with varying input meteorological data

library(GRAPLEr)
library(httr)
library(RCurl)
library(jsonlite)

# setup file system
# basedir <- ""
MyExpRootDir <- file.path(basedir, 'GRAPLE','MyExpRoot') 
dir.create(MyExpRootDir, showWarnings = FALSE)  
MyResultsDir <- file.path(basedir, 'GRAPLE','MyResults') 
dir.create(MyResultsDir, showWarnings = FALSE)
invisible(file.copy("job_desc.json", MyExpRootDir))

# copy files to Experiment Root Directory
run_example_sim("GRAPLE/MyExpRoot")

graplerURL <- "https://graple.acis.ufl.edu"  # specify web service address for the GRAPLEr.

MyExp <- new("Graple", GWSURL=graplerURL, ExpRootDir=MyExpRootDir, ResultsDir=MyResultsDir,
                ExpName="EDDIE", TempDir = tempdir())
MyExp <- GrapleCheckService(MyExp)

MyExp <- GrapleRunSweepExperiment(MyExp)
print(MyExp@StatusMsg)

MyExp <- GrapleCheckExperimentCompletion(MyExp)
print(MyExp@StatusMsg)