Quantifying the influx of water and elements to
ecosystems via atmospheric deposition is uncertain mainly because of
spatial variability; interpolation between precipitation stations is a
major source of uncertainty at the ecosystem scale. Various methods of
interpolation are used in precipitation and atmospheric deposition
studies, but the uncertainty in the interpolation is rarely reported.
Temporal dynamics generally contribute less uncertainty to estimates of
deposition, because precipitation amounts are measured at short
intervals (15 minute steps or shorter) or are cumulative, giving good
estimates of rainfall amounts at a point.
Thursday, June 16, 2011
Uncertainty in predicting precipitation volume at the Hubbard Brook Experimental Forest: A model comparison
Thursday, June 2, 2011
Meeting on Statistical Issues in Forest Management held on 2-4 May, 2011
A meeting was held a few weeks ago in Québec City on statistics and
forest management, with uncertainty as a central theme in most of the 20
or so invited talks. Tim Gregoire presented a fascinating opening talk
about the unresolved issue of propagating jointly model uncertainty and
sampling design uncertainty. Ron McRoberts presented a talk about
variance estimation within a k-NN imputation exercise. The meeting ended
on presentations by Bruce Borders and Annika Kangas on the opportunity
costs associated with uncertainties in forest inventories. The
presentation pdfs are on the meeting web site: http://www.crm.umontreal.ca/Forest11/horaire_e.php
- Courtesy of Pierre Bernier
Canadian Forest Service
Laurentian Forestry Centre
Quebec, Canada
- Courtesy of Pierre Bernier
Canadian Forest Service
Laurentian Forestry Centre
Quebec, Canada
Subscribe to:
Posts (Atom)