Skip to content

tarankalra1/observations_FI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FI_processing - Matlab scripts from Chris Sherwood to process Fire Island oceanographic data.


Description of folders

  • repres_orbital_vel - Code to compare ADV and workhorse based representative orbital velocity amplitude
  • recreate_time_series - Recreating time series from measured data (similar to Abreu et al.)
  • vandera_bedload - generated matfile that contains the vandera based bedload calculation (both from direct waveform and then from workhorse).
  • skewness - skewness is computed betweeen ADV calculations and empirical relationships.
  • compare_bedload - bedload calculations in vandera and measured values (MPM).
  • find_bedload_compare.m - code that compares the bedload from vandera and measured values - it does require matfiles that are first generated by the find_skewness_adv.m.
  • Steve_ana_waveforms- Steve generate the wave forms from analytical calculations
  • waveform_compare - compare the wave form characteristics empirical vs data from ADV(used Steve's mat file).

Overall goals of this repo

  • Compare asymmetry deduced from bulk wave statistics via Ursell number with asymmetry measure with near-bottom instruments
  • But first, demonstrate/test routines for reading/processing instrument data

Asymmetry

  • Measures of asymmetry vary. Some are best calculated for a singe wave, others are amenable to field measurements.
  • TODO: summarize asymmetry formulae and code

Progress so far

  • The stats file has some missing depths...we skip these in processing, but have not investigated the underlying bursts. Some of the bursts where depth is missing are also missing some pressure measurements. These break the calcs.
  • TODO: Add some basic QA/QC to the bursts to replace missing or bad data where possible, and to skip bursts that don't meet QA standards.
  • TODO: Complete analysis of rotation into wave-propagation direction. I think wds returns the correct direction, and puvq returns a direction that is 180 deg. out...the direction waves are coming from.
  • TODO: Complete calcs. of comparable statistics. Best version of du/dt for acceleration statistics? Hilbert transform (per Malarky and Davies, 2012) for acceleration asymmetry?

Code so far:

  • puv_proc_FI - Main script to read in ADV data and process burst by burst

  • puvq - Function to calculate wave parameters from burst measurements of u, v, p

  • ubstatsr - Function to calculate a range of burst statistics

  • ruessink_asymm - Function to calculate asymmetry parameters from Ursell number

  • urotate - Unfinished function to rotate velocities into direction of wave propagation

  • ustats - Unfinished function to calculate statistics on u returned by urotate

  • compare_advs - Reads and plots from both ADV statistics files.


Nortek scripts

We thank NortekUSA for several scripts written by Lee Gordon and distributed here:

http://www.nortekusa.com/usa/knowledge-center/table-of-contents/waves

These are:

wds.m - Directional wave spectra calcs

hs.m - Summary statistics of output of wds

wavek and logavg - called by wds

About

observational comparison of Fire Island

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published