Delete README.txt

####################################################################

#                                                                  #

#   ------------------------------------------------------------   #

#   A COMPARISON OF THE ATLANTIC AND PACIFIC BJERKNES FEEDBACKS:   #

#   SEASONALITY, SYMMETRY, AND STATIONARITY                        #

#   ------------------------------------------------------------   #

#                                                                  #

#   T. DIPPE, J. F. LÜBBECKE, R. J. GREATBATCH                     #

#                                                                  #

#   in                                                             #

#                                                                  #

#   JOURNAL OF GEOPHYSICAL RESEARCH: OCEANS                        #

#                                                                  #

#                                                                  #

#   Readme for using the Matlab scripts                            #

#                                                                  #

####################################################################

# --------- #

# TECHNICAL #

# --------- #

Matlab version used for all analysis: Matlab 2018

# ------------- #

# USED DATASETS #

# ------------- #

- AVISO-SSH:        https://www.aviso.altimetry.fr/en/home.html

- ERSST:            https://www.ncdc.noaa.gov/data-access/marineocean-data/extended-reconstructed-sea-surface-temperature-ersst-v5

- ERA-Interim USTR: https://www.ecmwf.int/en/forecasts/datasets/archive-datasets/reanalysis-datasets/era-interim

- ERA-40 USTR:      https://apps.ecmwf.int/datasets/data/era40-moda/

- ORAS4 data:       http://icdc.cen.uni-hamburg.de/projekte/easy-init/easy-init-ocean.html

# ------- #

# GENERAL #

# ------- #

# ANALYSIS USING A BUNDLE OF SCRIPTS

# ++++++++++++++++++++++++++++++++++

Many scripts will be made up of a group of three to four scripts. These are:

- <name>.m: The main script that does the principal calculations

- <name>_settings.m: Prescribing all kinds of settings, including paths and

  various plotting options

- <name>_plots.m or similar: Plots

- <name>_wrapper.m: A wrapper around the main script that allows looping over,

  for example, feedback elements or basins

For a given script <name>, I will refer to the bundle as <name>*.m

# ZONAL ANALYSIS VERSUS INDIVIDUAL INDICES

# ++++++++++++++++++++++++++++++++++++++++

All analyses are based on indices. However, I distinguish between "zonal

analyses" and "individual indices". Datasets for the zonal analyses contain a

matrix that holds time series of indices that correspond to a sliding index

along the equator. The zonal analysis is set up in

zonal_analysis_region_definition.m (for the published paper: 4 deg lon x

4 deg lat boxes that slide along the equator from a fixed western boundary

towards the eastern boundary), an the collection of sliding indices is

calculated with calculate_equatorial_indices*.m. These datasets should be stored

in the <path where data for zonal analysis is stored>.

In contrast, individual indices refer to just a single index. Single indices are

calculated with the function calcIndexAveraged() and should be stored in the

<path were individual indices are stored>.

Additionally, raw SST data might be required, such as ERSST. These datasets

should be stored in the <path where SST data is stored>.

# --------------- #

# LIST OF SCRIPTS #

# --------------- #

# FUNCTIONS

# +++++++++

- bootstrapFeedbackStrength() and bootstrapFeedbackStrengthDiff(): Significance

               test for feedback strengths and their differences.

- calcAnomalies(): Calculates the anomalies, and allows for variable detrending

- calcIndexAveraged(): Calculates an index, requires a file that describes the

                       grids that the variables are given on, including the

                       spatial extent of each grid cell to apply the correct

                       spatial averaging. (The grid files are stored in my case

                       in the file modelGrids.mat, which contains a structure

                       for each grid, which in turn stores information such as

                       box_grid_area.)

- calculateFeedback(): Perform a robust regression to estimate the strength of a

                       given feedback element.

- diagnoseLag(): Diagnose the lag at which the feedback strength between two

                 variables is maximum.

- eventCharacteristics(): Identifies events in a time series and returns various

                          statistical measures about them. This function is the

                          backbone of the SST event analysis in the Discussion.

- testValue(): Given a distribution, a significance level and a test value,

               discern whether the test value sits on the "significant" tails of

               the distribution.

# (More or less) PROJECT-WIDE SETTINGS

# ++++++++++++++++++++++++++++++++++++

- atlantic_asymmetric_bjf_plot_settings.m contains general settings

  for plots that will be called for a number of analysis scripts. Customize the

  colorbar handles in this script. However, the script is not used uniformly

  across all analysis scripts. Sorry for the inconvenience.

- zonal_analysis_region_definition.m defines the properties of the zonal

  analysis, i.e. the total zonal extent of the analysis region, as well as the

  width of zonal bins that will be used to calculate the equatorial indices.

# PREPARING THE DATA: Binning data into sliding indices along the equator

# ++++++++++++++++++

- calculate_equatorial_indices*.m: Calculate indices for each zonal bin defined

  in zonal_analysis_region_definition.m. These datasets are the basis for most

  of the analysis. They should be stored in the <path where data for zonal

  analysis is stored>.

Date vectors are expected to be Matlab date vectors (see

https://de.mathworks.com/help/matlab/ref/datevec.html).

# ANALYSIS

# ++++++++

Fig. 1: compare_oras4_ersst*.m

Fig. 2: diagnose_zonal_feedback_lags.*

<Fig. 3: Created in MacOS's Keynote>

Fig. 4-6: lagged_feedback_strength*.m to diagnose the strengths of the feedback

          elements, and diagnose_feedback_strength_difference_significance*.m to

          test the significance

Fig. 7: total_bjerknes_feedback_from_regression*.m

Fig. 8: compare_aviso_oras4_strengths_Atl3_Nino34*.m

Figs. 9-11: decadal_feedback_variations*.m

Fig. 12: zonal_enso_properties*.m

Fig. 13: decadal_sst_event_variations*.m

Figs. A1-A2: calculate_instantaneous_feedback*.m, and

             compare_lagged_instantaneous_feedback_strengths*.m

Figs. A3-A5: total_bjerknes_feedback_from_regression*.m

# ------- #

# CONTACT #

# ------- #

Contact me via tina.dippe@gmail.com if you have questions.