Modeling of water vapor transport through a dry soil layer (DSL), typically formed in the top soil during dry seasons in arid and semi-arid areas, is still problematic. Previous laboratory experiments in controlled environments showed that the only vapor transport process through the DSL is by Fick's law of diffusion. However, field experiments exhibited consistently higher evaporation rates than predicted by diffusion flow only. Some proposed reasons for the mismatch were: (a) daily cycles of condensation and evaporation in the DSL due to changes in solar radiation; (b) wind effects on air movement in the DSL; (c) atmospheric pressure fluctuations; (d) nonlinear influence of the DSL thickness on the evaporation process. To link the laboratory experiments with field observations, we performed soil column experiments in the laboratory with thick (>50 cm) DSL, and with different wind speeds, two radiative lamp schedules (continuous and 12 h daily cycles) and different thicknesses of DSL. Atmospheric pressure, air temperature and humidity were measured continuously. The results show that the evaporation rates observed are larger than those predicted by diffusion flow only. We found that it was possible to model the evaporation rates as a function of atmospheric pressure fluctuations. In conclusion, atmospheric pressure fluctuations can induce evaporation rates in DSL larger than estimated by diffusion flow only, possibly explaining the discrepancy between laboratory and field evaporation rates.