Premium
A Near‐Real‐Time Version of the Cross‐Calibrated Multiplatform (CCMP) Ocean Surface Wind Velocity Data Set
Author(s) -
Mears Carl A.,
Scott Joel,
Wentz Frank J.,
Ricciardulli Lucrezia,
Leidner S. Mark,
Hoffman Ross,
Atlas Robert
Publication year - 2019
Publication title -
journal of geophysical research: oceans
Language(s) - English
Resource type - Journals
eISSN - 2169-9291
pISSN - 2169-9275
DOI - 10.1029/2019jc015367
Subject(s) - environmental science , wind speed , data set , satellite , meteorology , remote sensing , computer science , geology , aerospace engineering , physics , engineering , artificial intelligence
The Cross‐Calibrated Multiplatform (CCMP) ocean surface wind data set was originally developed by Atlas and coworkers to blend cross‐calibrated satellite winds, in situ data, and wind analyses from numerical weather prediction. CCMP uses a variational analysis method to smoothly blend these data sources into a gap‐free gridded wind estimate every 6 hr. CCMP version 2.0 is currently produced by Remote Sensing Systems using consistently cross‐calibrated satellite winds, in situ data from moored buoys, and background winds from the ERA‐Interim reanalysis. The reanalysis fields are only available after a delay of several months, making it impossible to produce CCMP 2.0 in near real time. Measurements from in situ sources such as moored buoys are also often delayed. To overcome these obstacles and produce a near‐real‐time (NRT) version of CCMP (CCMP‐NRT), two changes are made to the input data sets: The background winds are now the operational 0.25‐degree NCEP analysis winds, and no in situ data are used. This allows CCMP‐NRT to be routinely processed with a latency of less than 48 hr. An intercomparison of the CCMP‐NRT results with CCMP 2.0, and independent measurements from moored buoys shows that CCMP‐NRT provides a modest improvement over the background wind from NCEP in regions where satellite data are available. Analysis shows that the inclusion of in situ measurement in CCMP improves the agreement with these measurements, artificially reducing estimates of the error.