Laserfiche WebLink
<br />MARCH 1980 <br /> <br />NOTES <br /> <br />336 <br /> <br />TABLE 2. Cross correlation table of predictor variables. Circled values are correlations greater than 0.49 <br />between variables which were not grouped in the same category. <br /> <br />< <br /> <br />/"~~I <br /> <br />~ <br />,,'" ,(., <br />~ ,,<:,'" <br /> <br />~ <br />~' ~o <br />'v 'v <br />" " <br /> <br />~ <br />~'" <br /><t <br /> <br />~ <br />~., <br /><t <br /> <br />... <br />..." <br /><t~ ~<>c <br /> <br />~ <br />~.. <br /><t <br /> <br />T SFC 11.00, .59 .57 <br /> <br />EV'20 1.00 <br /> <br />.29 -.02 .21 <br /> <br />.15 <br /> <br />.15 <br /> <br />.1 ~ <br /> <br />.18 -.21 <br /> <br />.85 G .2~ .25 .~7 .45 .4~ .41 <br /> <br />G .27 .~4 .~7 .47 <br /> <br />-.09 '.05 .08 <br /> <br />LEVo 1.00 <br /> <br />.47 <br /> <br />.46 -.21 -.1 5 .2~ <br /> <br />/ <br /> <br />~ ~ ~ ~ <br />~1Ioj Q:-ft) Q:-'" Q:-ltl <br />j' i' j' j' <br /> <br />~ ~ ~ <br />~"" c::J" c::J't) <br />~ ~ ~ <br />~ ~ ~ <br />:":\ <br />.12 .05 -.~~ -.6j <br /> <br />.05 .00 -.~9 8 <br /> <br /><>c ~ <br />.i <t<:'''' <br />~'" ~'" <br /> <br />~ <br />....' <br />" <br /> <br />~ A..,..... <br />:: A... ~ <br />-.29 8 .2~ .57 <br /> <br />e .49 <br />.12 .65 G .29 <br /> <br />-.42 .57 <br /> <br />-.4~ .60 <br />.../ <br /> <br />.10 .71 <br /> <br />.42 .26 -.40 .60 .12 <br /> <br />.20 -.11 -.40 <br /> <br />EPT,OO 1.00 .33 .~2 .48 .70 .74 .52 -.14 -.06 ,02 .30 .~7 .14 .10 -.44 G .06 .02 -.~2 -.~I <br /> <br />TOSFC 1.00 .64 .74 .70 .68 .8~ -.73 -.71 .46 .53 -.15 -.09 -.0788 -.20 .10 -.06 -.19 <br /> <br />PW.,o 1.00 .61 .62 .62 .72 -.73 -.68 .63 .47 -.06 .09 .0 I G G .08 .14 .05 -.12 <br /> <br />PW700 1.00 ,91 .89 .92 -.75 -.70 .50 .71 -.12 .02 .04 86 -.08 .O~ -.20 -.17 <br /> <br />PW,oo 1.00 .99 .87 -.62 -.55 .39 .75 -.04 .02 -.10 88 -.10 -.06 -.21 -.14 <br /> <br />PWTOT 1.00 .85 -.60 -.52 .36 .73 -.04 -.06 -.11 GG -.08 -.05 -.19 -.12 <br /> <br />MR 1.00 -.85 -,80 .61 .65 -.09 .01 .02 88 -.07 .17 -.23 -.22 <br />:'\ <br />LIIOO 1.00 .97 -.90 -.12 .24 .09 .05 .5':; -.40 -.02 -.28 .O~ .04 <br /> <br />LI,o 1.00 -.91 -.71 .24 -.01 -.09 G -.42 -.01 -.25 .07 .18 <br /> <br />TT! 1.00 .65 -.08 .07 .10 -.44 .39 .17 .33 -.01 -.12 <br /> <br />K I 1.00 -.28 -.24 -.22 G .41 .11 -.14 -.11 .02 <br /> <br />WOIR300 1.00 .59 .37 -.05 .36 .06 .19 -.17 -.46 <br /> <br />ficient was 0.73 and occurred between maximum <br />5 min rain volume and 70 kPa wind speed. In nine of <br />the dependent variable classes, the wind speed term <br />had the highest individual correlation coefficient <br />seven times and was second highest in the other <br />two. However, it was so insignificant in the maxi- <br />mum echo height Class that it was not entered into <br />the regression equation. <br />The water vapor-stability category also correlated <br />fairly well. It had the highest individual correlation <br />coefficient in three of the classes and the second <br />highest in four others. The stability index variable <br />emerged as the most highly correlated in this cate- <br />gory for six of these seven occurrences. <br /> <br /> <br />WOIR,OO 1.00 .69 -.08 .26 .32 .21 -.29 -.75 <br /> <br />WOIR700 1.00 -.07 .19 .23 .21 -.28 -.85 <br /> <br />WOIR.,O 1.00 -.95 .22 .20 .30 .38 <br /> <br />WSHEAR <br /> <br />Ir~ <br />1.00 -.18 -.13 -.33 1'\.:5~ <br /> <br />WSP0300 1.00 .73 .30 .00 <br /> <br />WSPD,OO 1.00 .21 -.07 <br /> <br />WSP0700 1.00 .49 <br /> <br />WSP0150 1.00 <br />'---- <br /> <br />The multiple correlation coefficients (Mult R) <br />ranged from 0.76 to 0.83 for all cases except the <br />maximum reflectivity class which had a value of <br />0.62. Mult R2 is the variance explained by the model <br />being tested and ranges from 58 to 69% except for <br />maximum reflectivity in which the explained vari- <br />ance was 38%. The dominance of the wind speed <br />and water vapor-stability terms shows up in the mul- <br />tiple regression analyses also. The wind speed and <br />water vapor-stability terms were the first two to be <br />entered into the equation during the stepwise pro- <br />cedure for eight cases (exceptions: maximum echo <br />height and maximum total area exceeding 10 dBz). <br />The average improvement to Mult R for these eight <br /> <br />