Reactive hydrocarbon flux footprints during canopy senescence

A coupled Lagrangian random walk and atmospheric turbulence model was employed to investigate the magnitude of isoprene source distribution within a mixed deciduous forest canopy undergoing defoliation. Modeled source distributions were studied to understand how the flux footprint evolved as the total amount and vertical distribution of foliage changed during the leaf senescing and abscission period. The modeled ensemble air parcel residence times inside the forest canopy were also studied to quantify the fraction of isoprene destroyed inside forest canopies due to rapid chemical reactions. Defoliation in the canopy affected the footprint by vertically redistributing the flux sources, and by reducing the leaf drag area encountered by flows within the canopy. For air parcel releases in the upper canopy, the increased in-canopy turbulence associated with defoliation shifted the footprint peak probability closer to the measurement point. However, when integrated through the depth of the canopy, the net effect of defoliation was to increase the upwind source areas farther from the flux measurement point. Defoliation also impacted air parcel residence times within the canopy. Under fully foliated conditions, air parcels remained within the canopy for periods ranging from 2 to 50 min, depending on levels of atmospheric turbulence and air parcel release height. Under 25–75% defoliated conditions, air parcels remained in the forest canopy for periods lasting less than 10 min. The estimated air parcel residence times inside the fully leafed canopy resulted in significant isoprene chemical processing. Based on Lagrangian footprint simulations with active chemistry, the integrated rates of isoprene destruction from reactions with ozone, hydroxyl, and nitrate radicals ranged from 12% for air parcels released in the upper canopy to 40% for air parcels released from the lower canopy. We conclude that active scalar flux estimates, often based only on the footprint transfer function and source strength distribution, can be substantially improved by incorporating an active chemistry term. # 2004 Elsevier B.V. All rights reserved.