Search Results

You are looking at 91 - 100 of 715 items for :

  • Regression analysis x
  • Weather and Forecasting x
  • Refine by Access: All Content x
Clear All
Brian F. Owens and Christopher W. Landsea

, that is, the differences between observed and predicted activity were aggregated into unit frequency intervals; root-mean-square error (rmse) analysis—the rmse of the annual differences between observed and predicted activity was compared; and regression analysis—observed tropical cyclone activity was regressed against that forecast using each of persistence, the statistical forecast, and the adjusted forecast to determine which forecasting method explained the greatest amount of variance in

Full access
Paul W. Mielke Jr., Kenneth J. Berry, Christopher W. Landsea, and William M. Gray

values indicate theexpected reduction in fit of the y and y- values for futureresults. Any tabled C4/ C1 value greater than 1.0 iscause for concern since this indicates that the sampleestimates of the population regression coefficients pro-vide a better validation fit, on average, than would bepossible had the actual population been available andis evidence that some sort of inflation of expected skillis present in the analysis. b. Population 1 Population 1 is the initial population of N

Full access
Frank Woodcock and Diana J. M. Greenslade

operational OCF where depleted component sets and up to half the observations in the learning window are tolerated before OCF is discontinued until sufficient observations are available for learning to resume. The restriction was imposed on the assumption that the more complex regression methods of bias correction and compositing could perform poorly compared to OCF when data were missing from the training period. The results comparing performance with varying training periods ( section 4e ) justifies

Full access
Jay S. Hobgood

analysis periods are summarized in Table 2 . The most active year was 1992, which produced 24 named tropical cyclones. There were also 20 or more named tropical cyclones in 1983, 1985, and 1990. The three least active years were 1988, 1991, and 1993. The summary statistics for the two periods chosen for the regression analyses are presented in Table 3 . There were nearly two more named tropical cyclones per year between 1982 and 1987 than there were in the period from 1988 to 1993. However the mean

Full access
Robert L. Vislocky and George S. Young

deriving linear regression equationsfor each of six flow types to forecast the POP. Woodcock (1980) forecast maximum temperatures by firstidentifying the 50 best analogs to the current sea levelpressure analysis and then applying screening regression to the predictors taken from these analogous cases.Thus, a new regression equation was developed for eachforecast. Results in both studies showed improvementover nonstratified equations. Despite the successes ofthese two models, two significant drawbacks

Full access
Robert L. Vislocky and J. Michael Fritsch

more responsive statistical guidance system. Preprints, l l th Conf. on Weather Forecasting and Analysis, Kansas City, MO, Amer. Meteor. Soc., 39-45.Cleveland, W. S., 1979: Robust locally weighted regression and smoothing scatterplots. J. Amer. Statist. Assoc., 74, 829-836.Dallavalle, J. P., J. B. Bower, V. J. Dagostaro, D. T. Miller, andJ. C. Su, 1992: Development of a new statistical weather forecastsystem. Preprints, 12th Conf on Probability and Statistics inthe Atmospheric Sciences

Full access
Katrina A. McDonnell and Neil J. Holbrook

, and whether forecasting on smaller spatial and temporal scales is likely to provide useful additional information for improving the skill of forecasting tropical cyclogenesis in the Australian– southwest Pacific region. This paper is structured as follows. The data are described in section 2 . The modeling methods and analysis techniques are explained in section 3 . Section 4 describes the Poisson regression models. The results are summarized and discussed in section 5 . Finally, section 6

Full access
Jung-Sun Im, Keith Brill, and Edwin Danaher

) are shown to exist and substantiate the use of regression model equations. This approach uses the linear regression of the AE on SP not only to predict the AE but also to derive a confidence interval (CI) for the AE from its distribution about the regression line. Using the regression model equation parameters derived at each horizontal grid point for each season and individual forecast lead time, we predict an AE associated with an individual SP and a 95% CI of the AE. Based on the AE CI forecast

Full access
Christopher A. Hiemstra, Glen E. Liston, Roger A. Pielke Sr., Daniel L. Birkenheuer, and Steven C. Albers

-station distance from nearest LAPS-assimilated data, and land cover). With the exception of land cover, site characteristics were regressed against temperature, relative humidity, wind speed, and precipitation measures (i.e., r   2 comparisons, rmse, and daily range r   2 values) using simple linear regressions. To evaluate the role of land cover, one-way analysis of variance (ANOVA) was performed using the temperature, relative humidity, wind speed, and precipitation estimates of variance ( r   2 ) as the

Full access
Robert J. Kuligowski and Ana P. Barros

predict the variables from real-time NWP model output. In addition to linear regression, a number of other data analysis techniques can be used to relate NWP model output to observed variables of interest. They include generalized additive models ( Vislocky and Fritsch 1995a ) and self-learning algorithms such as abductive machine learning ( Abdel-Aal and Elhadidy 1995 ) and goal-oriented pattern detection ( Dumais and Young 1995 ). Artificial neural networks are another approach that has been applied

Full access