Experiments in forecasting atmospheric marine horizontal visibility using model output statistics with conditional probabilities of discretized parameters.
Loading...
Authors
Karl, Michael L.
Subjects
model output statistics
visibility
North Pacific Ocean visibility
North Atlantic Ocean visibility
marine visibility
visibility forecasting
discretization
conditional probabilities
categorical forecasting
ocean areas
homogeneous ocean areas
thresholds
linear regression
natural regression
maximum probability
visibility
North Pacific Ocean visibility
North Atlantic Ocean visibility
marine visibility
visibility forecasting
discretization
conditional probabilities
categorical forecasting
ocean areas
homogeneous ocean areas
thresholds
linear regression
natural regression
maximum probability
Advisors
Renard, R.J.
Date of Issue
1984-06
Date
June 1984
Publisher
Monterey, California. Naval Postgraduate School
Language
en_US
Abstract
This report describes the development and application
of a program to forecast important air/ocean parameters using
the method (s) of model output statistics. The focus of this
operationally oriented study is to forecast atmospheric
marine horizontal visibility using a discrete analysis of
observed visibility and the Navy's Operational Global Atmospheric
Prediction System (NOGAPS) model output parameters.
Three strategies (two based on maximum-probability and one
based on natural-regression) are compared to two multiple
linear regression methods. The primary data set is from a
North Atlantic Ocean area bounded approximately by the North
American coast from Norfolk, Va. to St. Johns, Newfoundland,
and then eastward to about 37.5°W. Both the dependent and
independent data were derived from the same basic set. New
or unfamiliar concepts, in addition to the primary methodology,
include the statistical division of the North Atlantic Ocean
into physically homogeneous areas, two new threshold models
for the application of linear regression equations, linear
regression based upon a 'decision-tree' concept, functional
dependence of predictors and class errors. Results show
that the methodology proposed by Preisendorfer does out
perform multiple linear regression.
Type
Thesis
Description
Series/Report No
Department
Meteorology
Organization
Naval Postgraduate School (U.S.)
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Approved for public release; distribution is unlimited.
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.