To verify the accuracy of quantifying the soil water evaporation using the C–G model, we compared the actual evaporation losses with the calculated evaporation losses (Fig. 8). We observed that the calculated evaporation was close to the actual evaporation, and the variation trend followed by the soil water evaporation intensity with depth was approximately the same. During the entire test period, the relative error of the evaporation loss calculated based on δ18O was 13%, and the relative error of the evaporation loss calculated based on δ2H was 34%. The use of δ18O to quantify the soil water evaporation loss resulted in a 21% greater accuracy than that resulting from δ2H. However, the approach involving the use of isotopes had a lower evaporation loss, which may have been caused by the approximate treatment of the values of various parameters in the model calculation. Considering that there were some unavoidable sources of errors, such as the absolute error caused by the instrument itself, in field experiments, it is feasible to use stable hydrogen and oxygen isotopes combined with a modified C–G model to quantify the evaporation of bare soil under continuous-evaporation conditions.
Fig. 8 Comparison of the actual evaporation loss and calculated evaporation loss at different soil
depths and different periods from May 13, 2018 to June 27, 2019 (based on δ18O and δD
isotopes).