Other tools in the package goFlux

Aside from the typical workflow of flux calculation, one may wish to estimate other parameters from gas measurements, or modify the observation window of measurements.

Currently, this page only contains three complementary functions. To request more complementary functions, open an issue on GitHub or contact the maintainer of this package, Karelle Rheault (), directly.

crop.meas for additional modifications of the observation window

Crop measurements from the start, from the end, or adjust the deadband. Individual parameters can be provided for each measurement, in an auxiliary file, or the same parameters can be applied to all measurements.

Usage

Note

Code chunks under Usage sections are not part of the demonstration. They are meant to show you how to use the arguments in the function.

crop.meas(
  dataframe,
  auxfile = NULL,
  crop_start = 0, 
  crop_end = 0,
  deadband = 0, 
  max.obs.length = NULL,
  cham.close = "start.time", 
  cham.open = "end.time"
)

Arguments

dataframe a dataframe containing gas measurements, output from import or ID functions (click.peak2 or manID). Must contain the columns UniqueID, start.time, end.time and Etime.
auxfile data.frame; auxiliary data frame containing the columns UniqueID, crop_start, crop_end, deadband and/or max.obs.length.
crop_start numerical; number of observations to remove at the start of the measurement. To provide a unique value per UniqueID, the column crop_start must be present in the auxfile and the argument must be set to crop_start = "aux".
crop_end numerical; number of observations to remove at the end of the measurement. To provide a unique value per UniqueID, the column crop_end must be present in the auxfile and the argument must be set to crop_end = "aux".
deadband numerical; define a deadband at the start of measurements (seconds). To provide a unique value per UniqueID, the column deadband must be present in the auxfile and the argument must be set to deadband = "aux".
max.obs.length numerical; maximum length of measurement (seconds) to keep after cham.close. To provide a unique value per UniqueID, the column max.obs.length must be present in the auxfile and the argument must be set to max.obs.length = "aux".
cham.close character; if cham.close = "start.time" (default), the original start.time before cropping will be saved into a new column cham.close. Set cham.close = NULL to drop the original start.time.
cham.open character; if cham.open = "end.time" (default), the original end.time before cropping will be saved into a new column cham.open. Set cham.open = NULL to drop the original end.time.

Details

Note that crop_start and deadband “crop” the start of the measurements in the same way.

Note that if max.obs.length is defined, end.time is calculated based on cham.close (original start.time in dataframe, if cham.close = "start.time") and max.obs.length, regardless of other defined values in the function. If cham.close = NULL, end.time is calculated based on start.time and max.obs.length.

If cham.close and cham.open are originally present in dataframe, set arguments cham.close and cham.open to NULL.

Note that the arguments crop_start and crop_end accept negative values and can be used to include more data points before and after the previously defined start.time and end.time.

Examples

# Load example data
data(manID.G2201i)

# Create an auxfile to define different crop_start and crop_end per UniqueID
auxfile <- cbind.data.frame(UniqueID = c("meas1", "meas2"),
                            crop_start = c(2,10),
                            crop_end = c(5,0))

# The arguments "crop_start" and "crop_end" must be set to "aux" to use
# unique values from the auxfile
crop.G2201i <- crop.meas(dataframe = manID.G2201i, auxfile = auxfile,
                         crop_start = "aux", crop_end = "aux")

# Alternatively, cropping can be applied equally across all measurements
crop.G2201i <- crop.meas(dataframe = manID.G2201i, auxfile = auxfile,
                         crop_start = 5, crop_end = 5)

auto.deadband for an automatic selection of deadband per measurement

Loops through multiple gas measurements and automatically define a suitable deadband per measurement.

Usage

auto.deadband(
  dataframe,
  gastype,
  db.min = 0,
  db.max = 60,
  obs.win = 90,
  mov.win = 10,
  min.r2 = 0
)

Arguments

dataframe data.frame; output from import or ID functions.
gastype character string; specifies which gas column should be used to detect inflection point. Must be one of the following: “CO2dry_ppm”, “CH4dry_ppb”, “COdry_ppb”, “N2Odry_ppb”, “NH3dry_ppb”, “NOdry_ppb”, “NO2dry_ppb” or “H2O_ppm”.
db.min numerical value (seconds); minimal deadband per measurement.
db.max numerical value (seconds); maximal deadband per measurement.
obs.win numerical value (seconds); maximal observation window to include in the linear models. If possible, keept at least 30 seconds between db.max and obs.win for more accurate results.
mov.win numerical value (seconds); time increment between obs.win, from chamber closure to db.max.
min.r2 numerical value; minimum increase in R2 required to accept dropping more data points.

Details

For each measurement, a sequence of observation window is defined from chamber closure (+ mov.win) to obs.win. For each observation window, a linear model is calculated, from which the R2 is extracted. The ‘best’ deadband is then selected at the highest R2.

In db.min and db.max, negative values are accepted. This is especially applicable in cases where start.time already includes a deadband after chamber closure, and cham.close is not provided.

Note that, obs.win should represent the most linear part of the measurement, at the start of the measurement. For long chamber enclosures (more than a few minutes) select a longer obs.win more representative of your measurements. The default value of obs.win = 90s is based on previous observations that the most linear part of a measurement is often found within a short period (<2 minutes) at the start of the measurement, for small chamber sizes (~4L) (Johannesson et al. 2024). Similarly, the argument mov.win could be increased for very long chamber enclosures. Using a short mov.win (e.g., 1s) increases computational time, and is not recommended for very large data sets. The default value of db.max = 60 is based on dark chamber measurements on, e.g., soils. For Reco, a longer db.max (e.g., 90 seconds) is recommended to take into account a recovery phase of residual photosynthesis at the start of the measurement.

min.r2 can be used to restrict longer deadbands (or minimize dropping data points). For example, if an R2 of 0.990 is found with a deadband of 30 seconds, and an R2 of 0.992 with a deadband of 40 seconds (considering mov.win = 10), the best deadband is defined at 30 seconds if min.r2 = 0.005. By default, the highest R2 selects the best deadband (min.r2 = 0), but this may result in unnecessary loss of data points.

In gastype, the gas species listed are the ones for which this package has been adapted. Please write to the maintainer of this package for adaptation of additional gases.

Examples

# Load example data
data("manID.UGGA")

# Automatically select a deadband per UniqueID
auto_db <- auto.deadband(manID.UGGA, "CO2dry_ppm")

# Apply changes to dataframe using crop.meas
crop_UGGA <- crop.meas(manID.UGGA, auto_db, deadband = "aux")

iso.comp for isotopic composition determination

Extracts the y-intercept of the regression between the isotope ratios (13C, 15N or 18O) plotted againts the inverse of the gas concentration (CO2 or N2O).

Usage

iso.comp(
  dataframe,
  CO2dry_ppm = "CO2dry_ppm",
  delta_13C = "Delta_Raw_iCO2",
  N2Odry_ppb = "N2Odry_ppb",
  delta_15N = "d15N",
  delta_18O = "d18O",
  save.plots
)

Arguments

dataframe a data.frame containing measurements of isotopic fractions of CO2 or N2O and a UniqueID (or chamID) per measurements.
CO2dry_ppm, delta_13C character string; specifies which columns should be used for CO2 measurements, and the delta 13C, respectively.
N2Odry_ppb, delta_15N, delta_18O character string; specifies which columns should be used for N2O measurements, the delta 15N and the delta 18O, respectively.
save.plots character string; a file path with the extension .pdf to save Keeling plots. By default plots are not saved.

Example

data(manID.G2201i)
my.iso.comp <- iso.comp(manID.G2201i)

References

Johannesson, C-F., J. Nordén, H. Lange, H. Silvennoinen, and K. S. Larsen. 2024. “Optimizing the Closure Period for Improved Accuracy of Chamber-Based Greenhouse Gas Flux Estimates.” Agricultural and Forest Meteorology 359 (December): 110289. https://doi.org/10.1016/j.agrformet.2024.110289.