Boundary Layer Feedbacks at an Ecotone Forced by Climate Change: A Treeline Study

Why do some ecosystem properties and boundaries respond quickly and dramatically to recent climatic forcing, while others resist change? The likely answer lies in the interplay between the climatic forcing and the positive and negative feedbacks that delineate ecosystem functioning and boundaries. We study this interplay in the context of the climatically-constrained ecotone at alpine treeline. The sensitivity of a treeline to increasing temperature is likely modulated by the spatial structure of the moving edge. Diffuse treelines generally migrate upwards in elevation, even in the absence of clear regional warming, while abrupt treelines are much more resistant to change. We argue that a treeline engineers its own climate through interactions between spatial structure, airflow, and transport/transformation of sensible and latent heat, and responds to the engineered climate through spatially distributed changes in recruitment and survivorship of the trees. This dynamic feedback between the treeline's spatial structure and boundary-layer processes interacts with external climatic forcing to enhance or retard treeline movement. Here we present data on the contrasting patterns of airflow and atmospheric temperature distribution at an abrupt treeline and a diffuse treeline on Pikes Peak (Central Rocky Mountains, Colorado). Our previous research has established that both treelines are moving in response to 20th century warming; the diffuse treeline is moving approximately 5x faster and in the course of late 18th and early 20th centuries it has transformed from abrupt to diffuse morphology. At both sites the investigation was conducted during the growing season using a 9m moveable tower with temperature, relative humidity, and wind speed sensors distributed across 7 heights (abrupt) and 5 heights (diffuse). Data were taken as one minute averages, sorted by wind directions, and normalized by the value at the highest station. The data were visualized using spline interpolation in ARC GIS. The dominant air flow at the abrupt treeline site was (i)parallel to the treeline (19% of the time), (ii) uphill flow slightly askew to the treeline (10 degrees from parallel, 22%), uphill flow significantly askew to the treeline(approximately 40 degrees from parallel to treeline, 17%). During parallel flow the height of the 70% ambient flow isovel decreased smoothly (nearly exponentially) with increasing distance uphill from the treeline. It decreased from the height of 7.5m at the treeline to the height of 2.5m 16m uphill of the treeline, and the height of 1.5m 36m and 50m uphill of the treeline. The height of the 30% ambient airflow traced the same pattern, displaced approximately1-1.5m below the 70% ambient airflow isovel. The airflow pattern appeared to be dominated by the high friction of the treeline on the downhill side and the low friction of the tundra on the uphill side. The temperature profile during parallel flow was also “well behaved”, at night (there was no parallel flow during the day in our sampling window) the system showed a strong inversion throughout the treeline ecotone with a thicker layer of cold air closer to the treeline. This is likely due to nighttime drainage and damming, or simply due to lower mixing in the proximity of the treeline. The askew flow shows a strongly contrasting picture (the overall pattern was very similar for slightly askew and significantly askew flow). The 70% ambient airflow isovel remained relatively constant in height of 6-7.5m for the distance of 30 m uphill (twice the tree height, 2H), creating sheltered conditions. Uphill of the 2H it abruptly decreased in height bringing 70% ambient windspeed to 50cm above the ground for the remaining part of the uphill transect in the tundra. The height of the 30% isovel on the other hand decreased exponentially with distance from the treeline uphill. The overall picture is consistent with a stationary lee eddy creating sheltered conditions for 2H uphill of the treeline and accelerating outwash of the eddy, which brings very fast moving air in the proximity to the ground creating strong sheer stress and inhibiting seedling establishment. The temperature profile also supported the establishment of the leeward eddy. Daytime temperatures in the sheltered zone were relatively warm but not hot, owing probably to partial shading by the saplings present in the sheltered zone. At the end of the sheltered zone there was a vertical band of significantly colder air, corresponding to the flow over the eddy. This air was likely adjusted to the forest canopies below and thus significantly cooler than the air of the sheltered zone. There is some evidence of intrusion of this cooler air into the sheltered zone near the ground. Uphill of the 2H mark, in the tundra, the temperature profile shows a typical daytime pattern with strong heating hear the ground and much cooler temperatures aloft. The nightime temperature profile is also consistent with the formation of the leeward eddy. The sheltered zone contains cold air, and is punctuated at 2H by a vertical band of warm air that likely marks the uphill end of the eddy. This warm air is likely adjusted to the canopies of the forest below (our preliminary works shows that tree canopies at treeline store significant amounts of heat over night). At the diffuse treeline the predominant airflow was directly uphill to askew uphill. The vertical airflow profile is very dissimilar compared to the abrupt treeline. The diffuse treeline sets up a layer of fast moving air at approximately 1.5-3m height. This layer is significantly faster than the layer immediately below it and the layer immediately above it. The height of this fast moving layer increases from approximately 1.5m at the low portion of the treeline (where trees are approximately 7m tall) to the height of 3m in the uphill portion of the treeline where the trees are approximately 50cm tall. The uphill horizontal spatial extent of this layer is approximately 100-120m. Our preliminary evidence suggests that the fast moving layer prevents mixing of the layers above and below and may be capable of entraining significant amount of sensible heat and transporting it upslope.