contour_shift: Apply contour-shifting to bias correct
In IceCast: Apply Statistical Post-Processing to Improve Sea Ice Predictions

Description Usage Arguments Details Value References Examples

Apply contour-shifting to bias correct a predicted contour using existing mappings.

1
2
3

contour_shift(maps, predicted, bc_year, pred_start_year, reg_info, level,
  dat_type_pred, my_land_mat = land_mat, my_land = land,
  n_train_years = NULL)

`maps`	object obtained from the `create_mapping` function (see details)
`predicted`	array of predicted values of dimension year x month x longitude x latitude
`bc_year`	year to be bias-corrected
`pred_start_year`	year prediction array starts in
`reg_info`	a `reg_info` list (see documentation for `reg_info`)
`level`	concentration level for which to build contour
`dat_type_pred`	string indicating the format of the prediction: either "gfdl" or "simple" (see details)
`my_land_mat`	binary matrix specifying land locations
`my_land`	`SpatialPolygons` corresponding to the land
`n_train_years`	number of years prior to the current year used in fitting the bias correction

The map parameter is a list of length four that has the form of a list obtained from running the create_mapping function. The values start_year and endYear give the first and last year that were mapped. The variables obs_list and pred_list are lists of arrays with one 3-dimensional array for each region. The first dimension of each array coresponds to the year, the second dimension corresponds to the lines on which the region is being mapped, and the third dimension corresponds to the variables of interest. The first and second dimension are indexed sequentially. The variables for the third dimension are for the fixed points' x-coordinates, the fixed points' y-coordinates, the mapped points' x-coordinates, the mapped points' y-coordinates, the length of the mapping vectorsin the x-direction, the length of the vectors in the y-direction, and the angles of the mapping vectors. The predicted data array, predicted, should be a single array of dimension: years x longitude (304) x latitude (448). If dat_type_pred = ``simple"}, the values in the array should indicate whether each grid box is categorized to contain ice (1: ice-covered, 0: no ice, NA: land). If \code{dat_type_pred =``gfdl" the values in the predicted array correspond to the raw ice concentrations values predicted (including indicators for missing data, land etc.) formatted as in the CM2.5 Forecast-oriented Low-Ocean Resolution (FLOR) model produced by the National Oceanic and Atmospheric Administration’s Geophysical Fluid Dynamics Laboratory and converted to a Polar Stereographic grid (Vecchi et al. 2014; Msadek et al. 2014). Weights for converting to a polar stereograhic grid were obtained from the spherical coordinate remapping and interpolation package (SCRIP) (Jones 1997).

SpatialPolygons object of the adjusted region

Jones, P.W. "A user’s guide for SCRIP: A spherical coordinate remapping and interpolation package." Los Alamos National Laboratory, Los Alamos, NM (1997).

Msadek, R., et al. "Importance of initial conditions in seasonal predictions of Arctic sea ice extent." Geophysical Research Letters 41.14 (2014): 5208-5215.

Vecchi, Gabriel A., et al. "On the seasonal forecasting of regional tropical cyclone activity." Journal of Climate 27.21 (2014): 7994-8016.

## Not run: 
adj <- contour_shift(maps = discrep, predicted = emFeb2012, bc_year = 2012,
                     pred_start_year = 1980, reg_info, level = 15,
                     dat_type_pred = "gfdl")
plot(land, col = "grey", border = FALSE)
plot(adj, add = TRUE, col = "blue")

## End(Not run)