A coupled weather and wildland fire forecasting system with - PowerPoint PPT Presentation

A coupled weather and wildland fire forecasting system with assimilation of satellite remote sensing data Jan Mandel University of Colorado Denver and Institute for Informatics, Czech Academy of Sciences with Jonathan D. Beezley, Mary Ann Jenkins, Adam K. Kochanski, Volodymyr Y. Kondratenko, Lixin Lu, Sher Shranz, Martin Vejmelka Supported by NASA ROSES A.35 11-FIRES11-0038 and NSF DMS-1216481

!"#$!%&'()'*$#"+$,-(."+&%/01,(234546( 3(

!"#$!%&'()'*$#"+$,-(27%$+$,-8($-(/%9:(344;6( <(

344;(./,-/(=,/(>#$8( ?(

@A!B.!CA)(/8(/,(",-$D#/-$E(8F8-$+(

@A!B.!CA)(M1+*1,$,-8( NAAA(G1#$M/8-( '()*+,-./.!)*0.1!2%# A-.)8651!(/59+,*/(! )*0.1!2%#BA-.) "3/456.!58/! (.),./5(3/.>! /.15?;.! :.5(!590! #8/.! -3)808(@> ;5,*/! .)8++8*9+! /589 <3=.+ C+)*D.E 2890 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"#$%& #3.1!)*8+(3/.!)*0.1 "3/456.!7/.!+,/.50!)*0.1 L/-/(/88"+"%/01,( A=@.(G&$%(+1"8-&#$(8-/01,8( HCCA.IJKLC.(>#$(E$-$M01,(

A$*#$8$,-/01,(1G(-O$(>#$(/#$/( PF(/(%$Q$%(8$-(G&,M01,( • ! RO$(%$Q$%(8$-(G&,M01,("8(D"Q$,(1,(M$,-$#(,1E$8(1G(-O$(>#$(+$8O ( • ! C,-$#*1%/-$E(%",$/#%F:(*/#/%%$%(-1(-O$(+$8O(%",$8 ( • ! !"#$%",$(M1,,$M-8(-O$(*1",-8(SO$#$(-O$(",-$#*1%/-$E(Q/%&$8(/#$(T$#1(((

)Q1%Q",D(-O$(>#$%",$(PF(-O$( %$Q$%(8$-(+$-O1E( U$Q$%(8$-(G&,M01, L !"#$(/#$/W( L< 0 U$Q$%(8$-($V&/01,( A"DO-BO/,E(8"E$(X(4( ! (U$Q$%(8$-(G&,M01,(D1$8(E1S,( ! (>#$(/#$/(D#1S8(

RO$(>#$(+1E$%W(G&$%(M1,8&+*01,( G&$%( "D,"01,( 0+$( R"+$(M1,8-/,-(1G(G&$%W( <4(8$M(B(Y#/88(P&#,8(V&"MZ%F( 5444(8$M([(L$/E(\(E1S,(P#/,MO$82]?4^(E$M#$/8$(",(+/88(1Q$#(54(+",6(

RO$(>#$(+1E$%("8(#&,,",D(1,(/(>,$#(+$8O(-O/,(-O$(/-+18*O$#$(+1E$%(

@",E(",-$#*1%/01,( • ! .*#$/E(#/-$8(G1#(E"_$#$,-(G&$%8(E$*$,E(1,( !"#$%&'% ()"$*&)+,%-+".-'%."/+#%01%'-+%23+4%5)+% U",$/#(",-$#*1%/01,(1G(S",E(/8(/(G&,M01,(1G( • ! %1D2O$"DO-I#1&DO,$88(O$"DO-69( 67&8'%"2%'-+%!"#$% 9:;<4+%"=%+7&8'41%4;.&:"'-)"8 (2`&8-(%"Z$(*"$M$S"8$( %",$/#(",-$#*1%/01,("8($'/M-(G1#(%",$/#(G&,M01,86( ",E$*$,E$,-%F(1G(-O$(Q$#0M/%(+$8O(8*/M",D( wind speed CG(-O$#$(/#$(,1(@A!(,1E$8(&,E$#(a+:( • ! +/-O$+/0M/%%F($V&"Q/%$,-(-1(-O$(b)N=H)(S",E( #$E&M01,(G/M-1#89(( C-(D$-8(-#"MZF( W • ! R F m e s h l e v e l – ! RO$(O$"DO-8(1G(-O$(,1E$8(/#$(M1+*&-$E(G#1+(-O$( 2 D$1*1-$,0/%:(SO"MO("8(/(*/#-(1G(-O$(81%&01,( – ! RO$(D$1*1-$,0/%(Q/#"$8(/(%1-(,$/#(-O$(>#$( – ! RO$(/-+18*O$#"M(/,E(>#$(+$8O(O/Q$(E"_$#$,-(#$81%&01,8( – ! RO$(#$8&%-(E$*$,E8(1,((-O$(#1&DO,$88(%$,D-O9( – ! R/Z$(-O$(#1&DO,$88(%$,D-O(G#1+(U=cLd.)(1#(G&$%8e( midflame height WRF mesh level 1 roughness height

c=.=BcK==(g#1`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cd.( );"='3:+% I#&41="=%&#$% <(Z+(Y#"E8(( K"=94&1%H1='+)% =-+18*O$#"M(!1#$M/8-(L/-/( cK==(K*$#/01,/%(2!f5?6W(NAAA(7Kcd.(G1#$M/8-(/-( <Z+(5O#(#$81%&01,(( A/E/#(E/-/(( EAAFH%I85/+% /88"+ 9 %/-$E( <:+%$+'+85;#% (( G":+%R&#&.+)+#'%K+8"=";#%R&O"#.% P:;$38'%

@A!B.!CA)(d8",D(cK==h8(( N"DO(A$81%&01,(A/*"E(A$G#$8O(2NAAA6(J1E$%( :9*+$,96/03;)<$=)639<,$>?.94+@$ +6A+==+=$?94.9<$4.+$BCCC$ %DE'F$=)639<$>*+=@G$2)*$4.+$"#!"$ H3*5+*$%3<1)<$%)6I0+7$8*+J$$K.+$ ()3*,+*$L56$*+,)0/;)<$=)639<$9,$ I03(+=$,43;(3001$4)$()M+*$4.+$ ?.)0+$8*+$2)*+(3,4$3*+3J$D<+$)*$ 6)*+$,6300+*$3<=$8<+*$!56$ *+,)0/;)<$=)639<,G$I03(+=$ =1<369(3001$=+I+<=9<N$)<$4.+$8*+$ 0)(3;)<$I*)M9=+,O$ • ! P+*1$.9N.$*+,)0/;)<$0)(30$ ?+34.+*$2)*+(3,4$ • ! :9*+$,I*+3=$2)*+(3,4$ • ! F6)5+$2)*+(3,4$ • ! :/+0$6)9,4/*+$3<=$:9*+$*9,5$ 2)*+(3,4$$ ( 5<(

.+1Z$(",(@A!B.!CA)( =9(!&%%(MO$+"M/%(#$*#$8$,-/01,W( b9(g/88"Q$(-#/M$#W( =%P","(!&$%(7/-$D1#"$8(25<6( =%P","(!&$%(7/-$D1#"$8(25<6( JKLC.(U/,E(71Q$#(RF*$8W( JKLC.(U/,E(71Q$#(RF*$8W( • ! • ! J"'$E(!1#$8-( J"'$E(!1#$8-( • ! • ! .O#&P%/,E8( .O#&P%/,E8( • ! • ! Y#/88%/,E8( Y#/88%/,E8( !&$%(M1,8&+*01,(#/-$8( !&$%(M1,8&+*01,(#/-$8( RWXIFL% bigalk cJK7W( bigene co c10h16 !Ccc($+"88"1,(G/M-1#8( ch4 )+"88"1,(G/M-1#8(G1#(-#/M$#286( c2h4 h2 c3h5oh no no2 c2h6 )+"88"1,(1G(MO$+"M/%( )+"88"1,(1G(-#/M$#286( c3h6 so2 c3h8 nh3 8*$M"$8( ch3cooh p25 ch3oh oc1 cres @A!(EF,/+"M/%(M1#$( oc2 bc1 glyald @A!B7O$+( hyac bc2 isop g%&+$(#"8$:(-#/,8*1#-(/,E( macr mek .+1Z$(-#/,8*1#-:(*%&+$(#"8$( E"8*$#8"1,(1G(MO$+"M/%%F( ( mvk /,E(/"#(V&/%"-F(G1#$M/8-( */88"Q$(8+1Z$( tol (2gJ39i:(1T1,$j6(

J1E$%",D(1G($+"88"1,8( ( gJ(39i(M1,M$,-#/01,(>$%E(",(&DI+<(8"+&%/-$E(G1#(-O$(@"-MO(/,E( Y&$`"-1(>#$8(",(7/%"G1#,"/(&8",D(@A!B.!CA)B7N)J9( 5i(

!"#$(g#1*/D/01,(L$*$,E8( 1,(!&$%(J1"8-&#$( !&$%(+1"8-&#$(M1,-$,-( • ! 2!J76(k(ZD(S/-$#(*$#(ZD(1G( 1Q$,BE#F(G&$%(( =P1&-(495(1#(%$88(",(-O$( • ! @$8-$#,(d.(>#$(8$/81,( )'0,M01,(!J7(k(>#$( • ! *#1*/D/01,(8*$$E(E#1*8( -1(T$#1(( !&$%(/**#1/MO$8( • ! $V&"%"P#"&+(!J7(S"-O(/( 0+$(%/D($80+/-$E(G#1+( S11E(8"T$W(G#1+(5O(G1#(Q$#F( >,$(P#/,MO$8:(,$$E%$8:(/,E( D#/88:(-1(5444O(G1#(P"D( %1D89( RO$($V&"%"P#"&+(Z$$*8( • ! MO/,D",D(S"-O(-O$( /-+18*O$#"M(M1,E"01,89( 5a(

A$+1-$(=&-1+/-$E( @$/-O$#(.-/01,( 7.i5aBUl(!&$%( J1"8-&#$(/,E( R$+*$#/-&#$(.$,81#( S"-O(!&$%(.0MZ8( 5;(

!&$%(+1"8-&#$(+1E$%( R"+$(%/D(+1E$%W !" ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! • ! m k W(G&$%(+1"8-&#$(M1,-$,-(G1#(-O$(G&$%(M%/88( k • ! E W($V&"%"P#"&+(2G#1+(/"#(-$+*$#/-&#$:(#$%/0Q$(O&+"E"-F:(/,E( *#$88&#$6( • ! T k W(0+$(%/D(25(O1&#:(54(O1&#8:(544(O1&#86(2G#1+(S11E(-O"MZ,$886( • ! C,(/(#/",:(-O$($V&"%"P#"&+( E (P$M1+$8(-O$(81/Z$E(8-/-$(2/**#1':( 344^(S/-$#6(/,E(-O$(0+$(%/D(E$*$,E8(1,(#/",(",-$,8"-F( • ! C,-$D#/-$E(PF(/,($'*1,$,0/%(+$-O1E:($'/M-(G1#(M1,8-/,-(M1$mM"$,-( /,E(8-/P%$(G1#(/#P"-#/#"%F(%/#D$(0+$(8-$*( • ! RO$(G&$%(+1"8-&#$(+1E$%(#&,8(1,(/(M1/#8$(+$8O(G1#(/%%(G&$%(M%/88$8( $Q$#FSO$#$:(-O$,(-O$(G&$%(M%/88$8(/#$(+"'$E(-1D$-O$#(1,(-O$(>,$(( >#$(+1E$%(+$8O(G1%%1S",D(-O$(G&$%(+/*(( 5n(

L/-/(=88"+"%/01,( • ! J1E$%(8-/-$(o(E/-/(kp("+*#1Q$E(8-/-$( • ! .-/080M/%(/**#1/MOW(8"+&%/-$( 9:;0&0"4"'1% $"=':"035;#%;2%'-+%);$+4%='&'+U% I 2 / 6( • ! L/-/(o(+$/8&#$+$,-($##1#($80+/-$($'*#$88$E(/8( $&'&%4"O+4"-;;$ W( I 2 / q = 6([(*#1P/P"%"-F(E$,8"-F(1G( -O$(+$/8&#$+$,-( = (G1#(/(D"Q$,(+1E$%(8-/-$r(( • ! !1#$M/8-W( I 2 2 / 6W( 0+2;:+ (-O$(E/-/("8(/88"+"%/-$E( • ! =,/%F8"8W( I 3 2 / 6W( &Y+: (-O$(E/-/("8(/88"+"%/-$E( • ! b/F$8(-O$1#$+W( I 3 2 / 6(k(M1,8-( I 2 / q = 6( I 2 2 / 6( ( 5s(

t/%+/,(>%-$#( forecast state distribution gaussian, mean u f covariance Q : p f ( u ) = const exp( ! || u ! u f || Q ! 1 2 /2) observation Hu + gaussian error with zero mean, covariance R : p ( u | d ) = const exp( ! || Hu ! d || R ! 1 2 /2) analysis = after the data is taken into account, by Bayes theorem p a ( u ) = const p ( u | d ) p f ( u ) = const exp( ! (|| Hu ! d || R ! 1 + || u ! u f || Q ! 1 2 ) / 2) 2 = const exp( ! || u ! u a || P ! 1 2 /2) analysis mean u a = u f ! K ( d ! Hu f ) covariance Q a = ( I ! KH ) Q where K = QH T ( HQH T + R ) ! 1 is the Kalman gain matrix 34(