Reply to comment by Finnigan on “On micrometeorological observations of surface-air exchange over tall vegetation”

Abstract

I thank Dr. Finnigan for his correspondence and his insightful analysis of the assumptions involved in Lee (1998). Before responding to his comment, it is useful to state the objective of Lee's analysis which is to examine the role of a non-zero mean vertical velocity or ow divergence/convergence in determination of surface-atmosphere exchange rates by the eddy covariance technique. Adequate upwind fetch with a horizontally homogeneous source distribution is assumed and (weakly) 2D air motions leading to ow convergence/divergence occur at length scales much larger than the tower footprint. Because this debate is motivated by a need for conservation of mass in surface-air exchange measurements, which is unfortunately not apparent in some ®eld studies, arguments will need to be both practical and theoretically sound. The main point of dispute is the assumption about horizontal advection (assumption `b'). This assumption and assumptions `a' and `c', critiqued by Finnigan, are made in almost all observational studies of surface-air exchange. This emphasizes the need to examine them critically. Lee (1998) separates total advection into horizontal and vertical (or mass ow) components and postulates that vertical advection is generally the dominant one at sites where the usual ux observational criteria are met. However, it is understood that horizontal advection is not always much smaller than vertical advection. For example, horizontal advection is not negligible near the boundary where there is an abrupt change in the surface source strength (poor fetch; e.g., Mahrt et al., 1994; Sun et al., 1998), or at places where a large horizontal gradient of the scalar concentration exists in response to heteorogeneous surface source distributions (Raupach et al., 1992). Finnigan uses thought experiments and the results of a linear analysis of neutrally strati®ed ow over hills (Raupach et al., 1992) to illustrate how the concentration ®eld might respond to ow convergence/divergence. The mechanisms he has identi®ed offer valuable guidance for future observational and modeling studies of the advection problem. There are, however, a number of points in need of clari®cation from the perspective of real atmospheric ows: (1) The experimental goal is to quantify the surface ux, w0c0 y