20.04.2013 Views

Documentation of the Evaluation of CALPUFF and Other Long ...

Documentation of the Evaluation of CALPUFF and Other Long ...

Documentation of the Evaluation of CALPUFF and Other Long ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Results foor<br />

<strong>the</strong> TS me etric are presented<br />

in Figgure<br />

C‐33. FFLEXPART<br />

haad<br />

<strong>the</strong> higheest<br />

TS statistics<br />

with a score<br />

<strong>of</strong> 25.8% %. HYSPLIT ( (22.7%), CAMMx<br />

(21.2%), <strong>and</strong> SCIPUFFF<br />

(18.8%) followed,<br />

<strong>and</strong><br />

<strong>CALPUFF</strong>F<br />

<strong>and</strong> CALGRID<br />

closed ouut<br />

with <strong>the</strong> wworst<br />

(lowesst)<br />

TS values <strong>of</strong> 10.3% annd<br />

8.8%<br />

respectivvely.<br />

Figure C‐ ‐33. Threat Score (TS) statistical<br />

peerformance<br />

mmetric<br />

for thhe<br />

six LRT mmodels<br />

<strong>and</strong><br />

CAPTEX RRelease<br />

5.<br />

Overall, t<strong>the</strong><br />

spatial performance<br />

p e for CTEX5 wwas<br />

relatively<br />

equal betwween<br />

FLEXPAART<br />

<strong>and</strong> CAMMx,<br />

with CAMMx<br />

having th he best perfoormance<br />

for <strong>the</strong> FMS <strong>and</strong><br />

POD statisstics<br />

<strong>and</strong> FLEEXPART<br />

having<br />

<strong>the</strong> best performanc ce for <strong>the</strong> FAAR<br />

<strong>and</strong> TS staatistics.<br />

<strong>CALPUFF</strong>,<br />

SCIPUUFF,<br />

<strong>and</strong> HYSSPLIT<br />

were<br />

generallyy<br />

comparable<br />

in <strong>the</strong>ir spatial<br />

performmance<br />

for CTTEX5,<br />

with SSCIPUFF<br />

showwing<br />

marginnally<br />

better sccores<br />

in all fo our <strong>of</strong> <strong>the</strong> sppatial<br />

metrics.<br />

CALGRID consistentlyy<br />

exhibited t<strong>the</strong><br />

poorest<br />

performaance<br />

across all four spattial<br />

metrics.<br />

C.5.2 GLLOBAL<br />

ANAL LYSIS OF MOODEL<br />

PERFORRMANCE<br />

Figure C‐ ‐34 displays <strong>the</strong> FOEX peerformance<br />

mmetrics<br />

for t<strong>the</strong><br />

six LRT mmodels.<br />

CALLPUFF<br />

had thhe<br />

best FOEEX<br />

score (clos sest to zero) ) with a value<br />

<strong>of</strong> ‐2.6%. The second best performming<br />

model<br />

using <strong>the</strong>e<br />

FOEX metric<br />

is CAMx (44.7%)<br />

followwed<br />

by HYSPLIT<br />

(‐9.2%) a<strong>and</strong><br />

CALGRIDD<br />

(‐11.9%).<br />

SCIPUFF <strong>and</strong> FLEXPART<br />

had <strong>the</strong> ppoorest<br />

FOEX<br />

scores witth<br />

values <strong>of</strong> 20.4% <strong>and</strong> ‐ 28.2%<br />

respectivvely.<br />

35

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!