These studies imply that the resolution of standard operational surface anal- 

 yses in the oceanic environment for these types of events is somewhere 

 between 2/3 and 1/4 that of the best available practice. The object here is not 

 to denigrate the operational analyses, which typically do not have the benefit 

 of the ship reports and buoy data incorporated in the research analyses. 

 Rather, it is to emphasize and substantiate the fact that the selection of events 

 based upon characteristic parameters can be biased by the analysis procedures 

 used to depict the event. Although this discussion has been primarily confined 

 to the single parameter of sea-level pressure, the pressure (and the pressure 

 tendency) parameter is significant in that it is highly correlated with the geo- 

 trophic windspeed (and the geotrophic windspeed tendency) (Weinstein and 

 Sanders 1989). Indeed, Weinstein and Sanders (1989) acknowledge that winds 

 are of more interest to the maritime conmiunity than pressure. The question 

 then arises, why not use wind data directly to aid in classification of events? 

 The answer, unfortunately, is that surface wind data in these events are scarce, 

 typically forcing wind estimates to be inferred from the pressure field. 



In addition to the minimum sea level pressure, size is another physical 

 measure of a storm event. Size is more difficult to quantify objectively than 

 minimum pressure because the definition of size is, to some extent, arbitrary. 

 Nielsen and Dole (1992) consider several different measures of size. These 

 include (a) distance from the cyclone center to the nearest high pressure center, 



(b) distance from the cyclone center to the nearest adjacent cyclone center, 



(c) distance from the cyclone center to the nearest col (saddle point) of sea 

 level pressure, or (d) the horizontal area encompassed by the largest closed 

 isobar about the cyclone. Definitions (a), (b), and (d) are rejected because of 

 difficulties caused by ambiguities when applied to some scenarios. Defini- 

 tion (c) was adopted and termed the RADIUS because it is effectively the 

 distance from the system center to the outermost closed isobar. 



Nielsen and Dole (1992) applied this size criterion to virtually all cyclones 

 that formed during and within the boundaries of the Genesis of Atlantic Lows 

 Experiment (GALE; 13 January - 16 March 1986). Any analyzed low that 

 appeared on two consecutive charts (3 hr or 6 hr depending upon geographic 

 area) was included in the study. For climatological documentation purposes, 

 an cyclones were classed as stationary if their terminal Oysis) position was 

 within 400 km of their genesis position; all others were classed as travelling 

 cyclones. Because the stationary cyclones tended to form and die over the 

 eastern continental United States, they are not relevant to the present study and 

 will not be discussed further. 



Recognizing the depiction problem caused by oceanic data scarcity pointed 

 out by Pauley and Bramer (1992), and others, Nielsen and Dole subjectively 

 re-analyzed the operational charts and included additional available data. They 

 found that the operational analyses often did not identify a developing surface 

 low until it was 2 or 3 mb below the surrounding field. This tendency caused 

 the operational analyses possibly to miss small systems in early stages of 

 development. The effect of these omissions, however, tended to be limited to 

 small, short-lived cyclones and did not significantly affect the results 



Chapter 2 Operational Analysis Depictions 



