Modelling chemistry in the nocturnal boundary layer above tropical rainforest and a generalised effective nocturnal ozone deposition velocity for sub-ppbv NOx conditions

Measurements of atmospheric composition have been made over a remote rainforest landscape. A box model has previously been demonstrated to model the observed daytime chemistry well. However the box model is unable to explain the nocturnal measurements of relatively high [NO] and [O3], but relatively low observed [NO2]. It is shown that a one-dimensional (1-D) column model with simple O3-NOx chemistry and a simple representation of vertical transport is able to explain the observed nocturnal concentrations and predict the likely vertical profiles of these species in the nocturnal boundary layer (NBL). Concentrations of tracers carried over from the end of the night can affect the atmospheric chemistry of the following day. To ascertain the anomaly introduced by using the box model to represent the NBL, vertically-averaged NBL concentrations at the end of the night are compared between the 1-D model and the box model. It is found that, under low to medium [NOx] conditions (NOx < 1 ppbv), a simple parametrisation can be used to modify the box model deposition velocity of ozone, in order to achieve good agreement between the box and 1-D models for these end-of-night concentrations of NOx and O3. This parametrisation would could also be used in global climate-chemistry models with limited vertical resolution near the surface. Box-model results for the following day differ substantially if this effective nocturnal deposition velocity for ozone is implemented; for instance, there is a 9% increase in the following days peak ozone concentration. However under medium to high [NOx] conditions (NOx > 1 ppbv), the effect on the chemistry due to the vertical distribution of the species means no box model can adequately represent chemistry in the NBL without modifying reaction rate coefficients.