Data Package Summary    View Full Metadata

Hemispherical photos were taken at 35 interior grid corners in each land use site’s 30 m x 50 m permanent plot on uniformly overcast days during summer 1998. A Canon A-1 35 mm camera with a 7.5 mm lens was mounted on a tripod, positioned so the film plane was 0.5 m above ground, oriented N-S with a compass, and leveled. Ektachrome 100 slide film was used. Two exposures were made for each location.

The slides were processed and digitized in jpeg format by Kodak, Inc. at a resolution of 1536 x 1024 dpi. The jpeg images were screened for nearground obstructions. Fifteen of the 35 photo locations per site were used in subsequent analysis.

Hemiview software (v. 2.0, Delta-T Devices Ltd., Cambridge, UK) was used to process the images and calculate indirect, direct, and global (= indirect + direct) sky factors for each sample location, plus the proportion of global sky factor represented by direct beam radiation. Several images spanning a wide range of canopy density were used to develop protocols for setting optimal thresholds in converting from color to black and white images. Each B/W image was then analyzed at three threshold levels, the optimum level plus 5 threshold units (~ 2.5%) above and below the optimum. The results were averaged across the three threshold levels to provide one set of sky factor data per sample location.

Above-canopy incident radiation was calculated daily for the June-September growing season based on (1) latitude and longitude, (2) average atmospheric transmission, and (3) monthly averages for fractions of the day in diffuse radiation due to cloud cover for the Harvard Forest area. Data for 2 and 3 above were provided by researchers at the State University of New York at Albany. Below-canopy radiation was estimated daily across half-hour intervals as the sum of diffuse and direct beam components. Sky factors were then calculated from the above- and below-canopy radiation values and averaged both monthly and across the growing season.