A New Method for Observing Interception Over an Old-Growth Forest in the Eastern Amazon Region
Czikowsky, University at Albany, State University of New York, firstname.lastname@example.org
Fitzjarrald, University at Albany, State University of New York, email@example.com
Sakai, University at Albany, State University of New York, firstname.lastname@example.org
Moraes, Universidade Federal de Santa Maria, email@example.com
Acevedo, Universidade Federal de Santa Maria, firstname.lastname@example.org
Medeiros, University at Albany, State University of New York, email@example.com
Conventional methods of estimating rainfall interception have yielded a wide range of results for tropical rain forests, with interception estimates ranging from 9 to 25% of total precipitation for the Amazon. There are now many more long-term eddy flux measurement sites than sites at which the individual forest water budget components (total precipitation, throughfall, and stemflow) are measured. It is worthwhile to pursue a method for estimating interception using eddy-covariance data from the flux sites.
We introduce and describe a new, alternate method for observing interception using eddy-covariance data that could be applied to other flux tower sites worldwide in varying forest types. The approach is to estimate the 'excess' evaporation that occurs following individual events, using baseline diurnal evaporation time series obtained from long time series of flux data. We examine nearly 400 cases during the period between April 2001 and July 2003 at the Km67 old-growth rain forest site of LBA-ECO in the Tapajós National Forest near Santarém, Brazil. A ceilometer was used to identify the exact starting and ending times of the precipitation as high temporal resolution rainfall data (beyond 30 minutes) were not available. We compare our interception estimates to those reported from conventional methods, as well as the interception calculated from commonly-used models, including the interception model of Gash. To show how the method is applicable to other forest types, we compare interception for the Amazon with data obtained in a boreal forest in Manitoba, where similar eddy-covariance and ceilometer measurements were made during BOREAS.