This is an experimental product.
For a detailed explanation see below.
All information contained within reflect data collected prior to the time of issusance.
|| Current Vector
The ARS is an autoregressive spectral forecast model.
It begins with the past five days of velocity observations from a TABS buoy.
It determines the spectral properties in the 40-hour filtered signal, and then projects that information forward in time.
The 40-hour filtered signal is used in order to remove the high frequencies and estimate the long-term, subtidal currents that transport material.
The ARS is autoregressive because it uses the past currents and it is spectral because it analyzes the currents for significant underlying periodic signals.
If the conditions seen in the past five days do not change over the next 24 to 48 hours, then this forecast is a sophisticated form of a persistence forecast.
The model has no information about future changes in external forces, in particular changes in the winds, and can be expected to do poorly when this happens.
The top panel shows the most recent forecast and is valid for the period noted in the title header. The forecast was issued at the beginning of the period.
The second panel shows the forecast issued 24 hours ago, if the ARS model was successful in running yesterday.
Because 24 hours of observations are now available to evaluate the forecast's skill, the rms error in the current speed and direction are noted.
The third panel shows the forecast issued 48 hours ago and the forecast skill.
The forecast currents are magenta colored, the TABS observation are blue, and any missing observations are interpolated and shown in red.