A comparison between direct and pan-derived measurements of the isotopic composition of atmospheric waters.

The stable isotopes of water can be used to examine and quantify the contribution to atmospheric moisture from evaporation, transpiration and surface-waters. However, obtaining extensive and ongoing time series data of the isotopic composition of atmospheric moisture has been difficult. Presented here is an alternate method using an isotope mass balance approach to estimate the isotopic composition of atmospheric moisture using water samples collected from class A evaporation pans. While this evaporation pan method does not provide the high-resolution time series data that can be obtained from an isotope analyser taking in-situ measurements of atmospheric moisture, the method is relatively simple and inexpensive to set-up and maintain.
In this preliminary investigation, a comparison between the isotopic composition of atmospheric moisture estimated from the evaporation pan method and in-situ measurements of the isotopic composition of water vapour using a Fourier Transform Infrared (FTIR) spectrometer deployed at the Lucas Heights weather station in New South Wales is undertaken. Through comparison of the two series of hydrogen isotope data, an assessment of the evaporation pan method can be made. Although there was some agreement between the isotopic composition of vapour measured by the FTIR spectrometer and the estimation for the atmospheric moisture (R2 = 0.49), the comparison is sensitive to climatic parameters that vary significantly within a 24- hour period such as the relative humidity of air and the air and pan temperatures. Inverting the model to use the FTIR spectrometer measurements at an hourly resolution improved the performance of the model (R2 = 0.57). However, this also revealed that the model produced more depleted values of the evaporation pan water isotopes than those observed. In contrast, there was a variable relationship between the modelled and observed isotope values of atmospheric moisture. These conflicting results will need to be resolved before the evaporation pan method is broadly applied in isotope hydrology.

Figure 3: Time series of estimated δL at hourly intervals (solid line), as a weekly moving average (thin broken line) and observed δL as determined by approximately weekly grab samples (circles).