In RespChamberProc: Processing Data from Respiration Chambers

library(knitr)
opts_chunk$set(out.extra='style="display:block; margin: auto"'
    #, fig.align="center"
    , fig.width=4.3, fig.height=3.2, dev.args=list(pointsize=10)
    , message=FALSE
    , results='hold'
    )
knit_hooks$set(spar = function(before, options, envir) {
    if (before){
        par( las=1 )                   #also y axis labels horizontal
        par(mar=c(2.0,3.3,0,0)+0.3 )  #margins
        par(tck=0.02 )                          #axe-tick length inside plots             
        par(mgp=c(1.1,0.2,0) )  #positioning of axis title, axis labels, axis
     }
})
# genVigs("singleSeries")

CO2 fluxes from Chamber CO2 concentrations

isDevelopMode <- TRUE
library(twDev)
setwd('..');loadPkg()

if(!exists("isDevelopMode")) library(RespChamberProc)
set.seed(0815)      # for reproducible results

Preparing the chamber data

First, the data is loaded. This is a typical data.frame as obtained from a logger file of a light chamber with photosynthetic plants in it, for a single chamber measurement cycle. For timestamps spaced about every second, the sensor readings from CO2 concentration in ppm, vater vapour in ppm, chamber temperaturein degree Celsius and air pressure, here in kPa are recorded.

data(chamberLoggerEx1s, package="RespChamberProc")
ds <- chamberLoggerEx1s
head(ds)
plot( CO2_Avg ~ TIMESTAMP, ds, ylab="CO2 (ppm)", xlab="time (Minute:Second)")

RespChamberProc demands the pressure in units Pascal, so we need to convert from kPa.

ds$Pa <- chamberLoggerEx1s$Pa * 1000  # convert kPa to Pa

RespChamberProc works with concentrations of CO2 per dry air. There was vater vapour in the chamber and CO2 concentrations are measured per wet air. We need to compute the concentrations per dry air.

conc <- ds$CO2_dry <- corrConcDilution(ds)

Computing the flux

The CO2-flux is the change of concentration per time, i.e. the slope of the concentration change. Usually, the changing concentration within the chamber over time decreases the slope voer time. Hence, RespChamberProc fits several functional forms to the CO2~time relationship and reports the intial slope of the best-fitting form.

resFit <- calcClosedChamberFlux(ds
        , colConc = "CO2_dry", colTime = "TIMESTAMP"    # colum names conc ~ timeInSeconds
        , colTemp = "TA_Avg", colPressure = "Pa"        # Temperature in K, Pressure in Pa
        , volume = 1, area = 1                          # chamber dimensions m3 and m2
)

calcClosedChamberFlux expects the time in seconds. Providing a vector of POSIXct is ok. Moreover, the dimensions of the chamber must be given, the inside chamber air volume in cubic meter, and the area (in square meter) of the exchange area between the chamber and the studied object. This area is usually the area of the open bottom side ofthe chamber.

If the chamber encloses an objects, set the area to 1 and interpret the flux in micromol/second instead of micromol/square meter/second.

In order to convert (change of) concentrations to amount of substance, the pressure (in Pa) and temperature (in Kelvin) inside the chamber are required in addition to volume in order to apply the idal gas law.
The flux is proportional to pressure. Hence, if pressure is not recorded with the logger file, one may compute with a average pressure of Pa=100*1000, and correct fluxes post-hoc when an esitmate of air pressure becomes available from meteorological data.

Inspecting the results

Its a one-row tibble, i.e. a kind of tabular data like a data.frame, with fitting statistics and fitted model objects. While all the first columns are numeric, column times holds a vector of times after lag-phase for which concentration predictions have benn made by the model object given in column model (see explanations on the graph below).

resFit

The most important information are the estimate of the flux, and its uncertainty, both in unit mumol/m2/s:

resFit[,c("flux","sdFlux")]

Or in gC/m2/day:

fluxInMuMol <- unlist(resFit[,c("flux","sdFlux")])
convert_mumolPers_to_gPerday(fluxInMuMol)

The results can be visualized in a plot:

plotResp(ds, resFit, label="Example Series")

The fitted form (solid line) approximates the change of measured concentration (points) over time. Usually there is a lag-time (dotted vertical line) between concentration change, and when the time when the signal reaches the sensor. The flux after the intial lag-time is reported at the top left together with its uncertainty in mumol/m2/s. The label can be provided to distinguish different measurement cycles.