tekknosol / glimmr

Compute gasfluxes with R

glimmr

Gas Fluxes and Dynamic Chamber Measurements

Overview

The glimmr package provides functions to convert high frequency concentration data obtained by (dynamic-) chamber measurements into gasfluxes. Data recorded with different devices can be processed automatically. The data is separated into chunks covering the single chamber applications and containing additional meta information (e.g. IDs, temperature…) provided by a metadata file.

Furthermore glimmr contains functions to calculate gasfluxes with the ‘Boundary Layer Equation’ (BLE).

To calculate gasfluxes from high frequency chamber measurements, glimmr offers a cusomizable system of gas analyzer definitions (analyzer()), as well as two preconfigured devices with read_, process_ and inspect_ functions:

• process_chamber() & inspect_chamber(): custom devices
• read_gasmet(), process_gasmet() & inspect_gasmet(): GASMET
• read_losgatos(), process_losgatos() & inspect_losgatos(): LosGatos

Installation

# The package is not released on CRAN yet.
# Development version from GitHub:
# install.packages("devtools")
devtools::install_github("tekknosol/glimmr")

Getting started

Calculate fluxes from device records and a meta file containing information about single measurements. Fluxes are reported in [mmol/m²/d] using linear model fitting and robust linear model fitting with robust::lmRob()

# read data
gasmet <- read_gasmet("path/to/gasmet_file.txt")
meta_gasmet <- read_csv("path/to/meta_file.csv")

# create diagnostic plot(s)
inspect_gasmet(gasmet, meta_gasmet)

# compute fluxes in mmol/m²/d
# F_LM: Liner Model
# F_RLM: Robust Liner Model
process_gasmet(gasmet, meta_gasmet)
#> End of interval determined by number of observations. Count = 10
#> Using temperature from meta file. Column = temp
#> # A tibble: 18 x 9
#>    date       site  begin               gas     rep     F_LM   LM_r2
#>    <date>     <chr> <dttm>              <chr> <dbl>    <dbl>   <dbl>
#>  1 2017-04-12 plotA 2017-04-12 09:01:49 CH4       1  0.0259  0.0119 
#>  2 2017-04-12 plotA 2017-04-12 09:08:56 CH4       2 -0.144   0.360  
#>  3 2017-04-12 plotA 2017-04-12 09:14:25 CH4       3  0.0980  0.118  
#>  4 2017-04-12 plotB 2017-04-12 16:47:41 CH4       1  0.0151  0.00160
#>  5 2017-04-12 plotB 2017-04-12 16:55:41 CH4       2 -0.0601  0.0449 
#>  6 2017-04-12 plotB 2017-04-12 17:02:20 CH4       3  0.107   0.307  
#>  7 2017-04-12 plotA 2017-04-12 09:01:49 CO2       1 -7.58    0.533  
#>  8 2017-04-12 plotA 2017-04-12 09:08:56 CO2       2  1.73    0.0987 
#>  9 2017-04-12 plotA 2017-04-12 09:14:25 CO2       3 -0.945   0.101  
#> 10 2017-04-12 plotB 2017-04-12 16:47:41 CO2       1 25.9     0.987  
#> 11 2017-04-12 plotB 2017-04-12 16:55:41 CO2       2 23.1     0.953  
#> 12 2017-04-12 plotB 2017-04-12 17:02:20 CO2       3 26.6     0.991  
#> 13 2017-04-12 plotA 2017-04-12 09:01:49 N2O       1 -0.0419  0.282  
#> 14 2017-04-12 plotA 2017-04-12 09:08:56 N2O       2  0.0262  0.279  
#> 15 2017-04-12 plotA 2017-04-12 09:14:25 N2O       3  0.0111  0.153  
#> 16 2017-04-12 plotB 2017-04-12 16:47:41 N2O       1  0.00409 0.0134 
#> 17 2017-04-12 plotB 2017-04-12 16:55:41 N2O       2  0.0154  0.176  
#> 18 2017-04-12 plotB 2017-04-12 17:02:20 N2O       3  0.0121  0.0712 
#> # … with 2 more variables: F_RLM <dbl>, RLM_r2 <dbl>

Calculate gas transfer coefficient used with the Boundary Layer Equation (BLE).

# Windpseed in m/s
windspeed <- c(2, 4.3, 1.8) 
# Water temperature in °C
watertemperature <- c(18, 21, 19.3) 

# Calculate gas transfer coefficient
calc_kW(windspeed, watertemperature)
#> [1] 1.488888 7.426978 1.246858

glimmr's Issues

no match between meta and gasmet file

Check for valid meta data (e.g. begin matches gasmet datetime) and throw error
Right now some internal R error appears that is not very helpfull

save diagnosis plots

save plots for flux-diagnosis instead of interactive mode

Fit non-linear function

Use non-linear function for flux calculation

use rlang::.data$xx in dplyr calls

To avoid notes in R CMD CHECK call dplyr etc. like:

meta %>% select(.data$CO2mmol)

prerequisite:
Import .data from rlang

throw error on unrecognized losgatos dates

don't produce empty df silently

Documentation

References

Vignettes

replace %>% pull(xx) structure

Use meta[[xx]] instead.

use ppm in inspect_()

use ppm instead of mmol

rename package

• rename .rproj file
• rename folder
• git remote set-url origin [GitHub repo URL]
• change value of package property in DESCRIPTION
• rename pkg-name.R
• edit README (title, ...)

check if columns exist

Fail if column definitions and columns in data and/or meta file do not match

gerneralization of preprocess_gasmet()

• Don't restrict to CO₂ AND CH₄
• use default windowsize if not provided
• use default meta information if not provided
• warn or throw error when air pressure is missing

use quotation for analyzer() parameter

allow analyzer(temperature = mean(t1, t2))

offset notation

0:0 notation written in excel causes csv file to include 00:00:00
Maybe use different separator

LosGatos temperature

use temperature:
start & end provided: mean
start provided: start
non provided: LosGatos ambient temperature

CRAN release v0.2.0

• Check that description is informative
• devtools::check_win_devel()
• rhub::check_for_cran()
• revdepcheck::revdep_check(num_workers = 4)
• Polish NEWS
• If new failures, update email.yml then revdepcheck::revdep_email_maintainers()
• Bump version (in DESCRIPTION and NEWS)
• devtools::check_win_devel() (again!)
• devtools::submit_cran()
• pkgdown::build_site()
• Approve email
• Tag release
• Bump dev version
• Write blog post
• Tweet

delete pics folder

delete files in pics folder. image override seems not to be reliable

Use atm. pressure from gasmet file instead of meta

calc mean atm. pressure from gasmet data for each flux

read LosGatos folder

function to read all files within specific LosGatos folder.
use ziped files and txt only when no zip is existing

Define custom devices

Define custom device to call functions like:

device <- analyzer(...)
process_chamber(data, meta, device)
inspect_chamber(data, meta, device)

• implement analyzer as S3 objects
• allow gals to be updated
• update documentation