ATL09&Layer&Parameters&
The$detection$of$layer$heights$will$be$performed$on$the$atmosph eric$ba cksc atte r$profiles$o f$each$of$the$
3$strong$beams$using$the$Density$Dimension$Algorithm$(DDA)$that$is$detailed$in$Part$II$of$the$
Atmospheric$ATBD.$Please$refer$to$that$document$for$information$pertaining$to$the$algorithm.$Here$we$
define$the$parameters$associated$with$the$layers$retrieved$by$the$DDA $and $their$rela ted$flag s.$Cloud$
detection$is$also$performed$using$an$estimate$of$the$Apparent$Surface$Reflectivity$(ASR)$for$each$strong$
beam.$This$method$does$not$retrieve$the$height$of$the$layer.$It$only$tells$whether$a$cloud$is$present$
somewhere$w ithin$the$atmosp heric$colum n.$It$is $as sumed$th a t$t h e $clo u d $d e te c tio n $fo r $th e $s tro n g $b e a ms$
represents$the $clou d$co nd ition $for$the $we ak $bea ms$at$the$same$geophysical$segment$(ATL03$
segment_id)$since$the$tracks$are$separated$by$only$90$m$and$½$second$in$time.$
There$will$be$a$maximum$of$six$layer$heights$(top$and$bottom)$stored$on$the$product$for$each$of$the$3$
strong$beams$at$25$Hz$resolution$–$ATL09$parameters$layer_top$and$layer _b o t.$$A ss oc ia ted $w ith $th e se $
will$be$a$layer$confidence$flag$(parameter$layer_co n f)$an d $a$laye r$attribu te$flag $(para m e ter$ lay e r_ a tt r).$
The$confidence$flag$is$calculated$by$averaging$the$bins$within$a$layer$and$dividing$it$by$the$average$
molecular$backscatter$for$the$same$vertical$range.$The$integer$value$of$this$calculation$will$constitute$
the$confiden ce$flag$(a $value$from$0$to$approximately$100).$The$layer$attribute$flag$is$intended$to$
discriminate$between$cloud$and$aerosol.$If$the$layer$bottom$is$above$6$km,$then$the$layer$is$cloud$
(layer_ a ttr$=$1).$If$the$layer$top$is$below$6$km$and$the$layer$confidence$is$less$than$10,$then$the$layer$is$
aerosol$(layer_attr$=$2).$
ATL09&Cloud&Flags&
In$addit io n $to $t h e$la y er $p ar a m e te rs$ ab ove,$there$are$two$cloud$flags$on$ATL09:$Cloud_Flag_Atm$and$
Cloud_Flag_ASR.$The$first$is$a$clou d$f lag $w h ic h $ca n $b e$u s ed $t o $de t ermine$wh e th e r$o r$n o t$a $lay e r$w a s$
detected$for$a$given$atmospheric$profile$(ATL09$product$parameter$Cloud_Flag_Atm).$This$flag$is$set$to$a$
positive$number$corresponding$to$the$number$of$layers$found$by$the$DDA$algorithm.$ATLAS$atmospheric$
data$will$have$a$much$higher$signal$to$noise$ratio$at$night$as$compared$to$day.$Detection$of$thin$clouds$
(optical$depth $<$1)$w ill$prob ab ly$no t$be $po ssib le$in$da yligh t$from $th e$ba cks catte r$pro files$at$full$
resolution$(25$H z).$Th e $ex ac t$limit$will$depe nd$on$the$solar$zenith$angle$and$surface$albedo.$Thicker$
clouds$should$be$detected,$but$many$of$these$will$be$so$thick$that$no$surface$signal$will$be$obtained.$
While&this&flag&can&be&used&during&the&day,&it&might&prove&better&to&use&the&ATL09&p rod uc t&par am e ter &
Cloud_Flag_ASR&which&is&based&on&Apparent&Surface&Reflectance&and&discussed&below.&But&during&
night,&Cloud_Flag_Atm&is&defin it e ly &th e &b e s t&c lo u d &fla g &to &u s e .$As$an$example$of$how$to$use$the$cloud$
flags$derived$from$the$backsca tter$profiles,$suppose$Cloud_Flag_Atm$is$equa l$to $2 $(in d ica tin g $th e $D D A $
found$2$layers ).$On e$w o uld $then $loo k$at$the $ la y er_ c o n f$parameter$to$see$if$any$of$those$two$layers$had$a$
confidence$above$say$50.$The$50$would$indicate$a$relatively$high$confidence$that$this$was$a$cloud$and $a$
fairly$optically$thick$clou d$a t$that.$Th e$co n fiden ce $leve l$50$us ed $here $is$sub ject$to $cha ng e$afte r$w e$loo k$at$
actual$ATLAS$data.$If$one$wanted$to$screen$all$clouds$from$the$data,$then$one $wo u ld$use $a$con fide nce $
level$of$10 .&