Estimate VPD from assuming dewpoint at daily minimum temperature

Source: R/estimate_vpd_from_dew.R

VPD is required for daytime NEE flux partitioning. Hence, it is necessary to estimate VPD also for long gaps in data. With two assumptions, VPD can be estimated from temperature 1). The change of water mass in air is negligible during the day. VPD is the difference of actual vapour pressure to saturation vapour pressure. 2.) At morning minimum temperature, vapour pressure is at minimum in many cases at saturation. Hence $$VPD = Esat(Tair) - E \approx Esat(Tair) - Esat_{daymin} \approx Esat(Tair) - Esat(Tair_{min})$$

estimate_vpd_from_dew(df, pNonMissing = 0.1)

Arguments

df

data.frame with columns DateTime, VPD, Tair, and Tair_f

pNonMissing

numeric scalar of the necessary fraction of finite VPD and Tair. If fraction is lower then a warning is thrown.

Value

numeric vector of length(nrow(data)) of estimated VPD

Details

Since sometimes Esat_daymin is lower than Esat(Tair_min) the estimated VPDfromDew is underestimated. This function applies a linear model of the existing VPD and estimated VPD to correct for this bias: VPD ~ 0 + VPDfromDew * Tair_f * hourOfDay * TminOftheDay * TRangeDay