Harvard Forest Symposium Abstract 2009

Title: A Cost-Effective Automated Sensor Network for Meteorological and Snow Depth Measurements
Primary Author: Robert Hellstrom (Bridgewater State University)
Abstract:
Cost and time requirements for consistently measuring meteorological forcing and snow depth in areas with heterogeneous physiography often limit the potential for snow model and remote sensing validation. This paper reports on the development and testing of an automatic sensor network (ASN) that measures soil and snowpack temperature profiles, snow depth, air temperature, humidity, reflected light intensity, and wind speed. The anemometer, consisting of half ping-pong balls attached to four arms extending orthogonally from the shaft of a low-resistance electric motor, was calibrated in a wind tunnel. Recent deployments of the ASN at mixed deciduous, mature Hemlock and open sites within the Harvard Forest of northern Massachusetts suggest low maintenance requirements under a broad range of meteorological conditions. Each ASN (Fig. 1) consists of three HOBO® loggers with internal and external sensors and one outdoor surveillance camera that takes daily snapshots of an 8X8 meter graduated snow stake grid (Fig. 2) and stores them on a laptop computer. Bi-weekly site visits provided snow pit density profiles of a snowpack (Fig. 3) with approximately 0.5 m depth at maximum. Comparisons of hourly measurements at the three sites show important differences between soil-snow temperature profiles and other measured variables that will improve the forest-snow algorithms in snow models.

Annual snowpack magnitude and dynamics influence the water retention and runoff characteristics of watersheds. Headwater streams and wetlands provide natural flood control, recharge groundwater, trap sediments, recycle nutrients, support biological diversity, and sustain the productivity of downstream rivers, lakes, and estuaries. Such streams also comprise at least 80% of the nation’s 17 stream network and offer the greatest opportunity for exchange between terrestrial and aquatic systems (Meyer et al. 2003). Despite providing essential ecosystem services that are increasingly threatened, headwater streams are relatively poorly studied and receive limited regulatory protection. In particular at HFR, intensive measurements of snowpack are lacking, which will contribute to a better understanding of wetland hydrology.

Furthermore, Hemlock forests, particularly in New England, USA, could functionally disappear in a few decades from the outbreak of Hemlock Woolly Adelgid (Adelges tsugae), a rapidly spreading Asian insect that kills hemlocks of all sizes within 4-15 years of infestation, and associated increases in hemlock harvesting. Although mortality is not strongly related to stand or site factors, the low damage in northern Massachusetts may be related to cold winter temperatures that reduce HWA populations (Paradis and Elkinton 2005).

Results from 2007-2008 of snow water equivalent (SWE) (Fig. 4) indicate approximately 25% greater SWE at both forest sites when compared to the open site with nearly identical peak SWE occurring at both sites after the last major snowstorm. Compared to the open site, melt-out was delayed by 3 days at the hardwood and 5 days at the Hemlock sites. Air temperature at the Hemlock site was higher than that of the open site for open site temperatures below -5 C, and it was lower than the open site for temperatures above -5 C (Fig. 5). This trend toward warmer temperatures below the Hemlock during colder conditions will not lead to significant impact on the snowpack, but the lower temperature when open site values are just above-freezing are plausibly responsible for the delayed melt-out at the Hemlock site. This may significantly deter the Wooly Adelgi infestation by keeping air temperature cooler for a longer period during the snow season, thereby prolonging the survival of the Eastern Hemlock. None of the sites experienced soil freezing at or below -10 cm depth (Fig. 6). Finally, and particularly at the hardwood site, several snow meltwater influx episodes were evident as day-long drops of 0.5 to 2 C in substrate temperature at -40 cm. This project will continue to monitor snowpack in the foreseeable future to improve understanding of and reveal the most important processes controlling snow cover in forested regions of central Massachusetts.
Research Category: Forest-Atmosphere Exchange