垂直干旱指数在湖北漳河灌区遥感旱情监测中的应用_英文_

垂直干旱指数在湖北漳河灌区遥感旱情监测中的应用_英文_ Pe rpendicula r D rought Index and its App lica tion in Hube i Zhanghe Irrigation System 1 1 2 1L I Zhe, TA N D e2ba o, Q I N Q i2m in g, ZHA NG Su i ( 1. Sp a tia l Info rm a tion Techno logy App lica tion In stitu te, Yangtze R ive r Sc ien tific R e sea rch In stitu te, W uhan 430010 , Ch ina; )2. Schoo l of Ea rth and Sp ace Sc ience, Pek ing U n ive rsity, B e ijing 100871, Ch ina A b stra c t: R emo te sen sing d rough t mon ito ring mode l is an impo rtan t re sea rch rea lm in irriga tion and wa te r con se rva tion. ( ) F rom a suppo sition tha t the sp ec tra l cha rac te ristic of ba re so il m ee ts linea r d istribu tion Zhan, 2005; A buduwa sit, 2004in () the N IR N ea r Infra red2R ed reflec tance sp ec tra l in fea tu re sp ace, a simp le and effec tive remo te sen sing mode l ha s been u sed to mon ito r d rough t wh ich wa s va lida ted in N ingxia H u i2na tion A u tonomou s region and ach ieved be tte r re su lts. In o rde r ( ) to ve rify fea sib ility of PD Ip e rp end icu la r d rough t indexunde r the den se vege ta tion cove r cond ition, H ube i Zhanghe Irriga2 tion System wa s se lec ted a s the re sea rch a rea. PD I wa s ca lcu la ted w ith ETM + da ta, and 2 - 5 cm so il mo istu re ob se rva2 tion re su lt wa s u sed in comp a rison w ith PD I. R e su lt show s tha t 2 - 5 cm in2situ ob se rva tion and PD I ca lcu la tion d isp lay ba2 sica lly the sam e trend and they a re con sisten t w ith the fie ld m ea su red d rough t ind ice s. To sum up , PD I can reflec t the land cove r and wa te r2hea t com b ina tion. PD I ha s m any advan tage s such a s simp lic ity, p rac tica lity, ea sy2to2u seage, ve rsa tility and un ive rsa lity. A nd it is su itab le to mon ito r regiona l d rough t unde r comp lex unde rlying su rface cond ition s. Key word s:N IR 2R ed Sp ec tra l Sp ace; Pe rp end icu la r D rough t Index; H ube i Zhanghe Irriga tion System CL C n um ber: P237; TP79 D ocum en t code:A tion index and wa te r stre ss index, such a s AV I and sp ec trum sp ace m e thod s. B u t becau se of d iffe ren t so il 1 INTROD UC T IO N struc tu re s, exte rio r shap e s and o rgan ic m echan ism s, the p rec ision of V isib le2N IR mode l to mon ito r d rough t R egiona l d rough t b rough t by globa l c lim a te chan2 is stric tly re stric ted. The rm a l mode l ca lcu la te s so il ge s is m enac ing d irec tly hum an’s food supp ly and su r2 mo istu re th rough so il su rface em issivity and land su r2 viva l environm en t, and it becom e s a ho t issue of m any face temp e ra tu re on the ba sis of so il wa te r ba lance e2 sc ien tists and gove rnm en t adm in istra tion s. R emo te qua tion and ene rgy ba lance equa tion, such a s the rm a l sen sing ha s m any advan tage s such a s la rge cove r, h igh ine rtia mode l, vege ta tion evapo ra tion mode l and wa te r sp a tia l re so lu tion, sho rt in te rva l, and conven ien t da ta ( ) dep re ssion index. W a tson e t a l. 1971 b rough t ou t a acqu isition , wh ich m ake s remo te sen sing d rough t mon i2 simp le the rm a l ine rtia equa tion w ith land su rface tem 2 to ring becom e an impo rtan t re sea rch rea lm. D rough t ( ) ( ) p e ra tu re. P rice 1977; 1985 , Kah le 1977 , England index is a comp rehen sive de sc rip to r of su rface d rough t ( ) ( ) ( ) ( ) 1990 , Zhang 1997 , Tian 1991 a lso deve lop ed and we t cond ition N a ra sim han & Srin iva san, 2005 . m any app lica tion exp e rim en ts. The rm a l ine rtia mode l Ow ing to d iffe ren t p hysica l and rad ian t cha rac te rs of so il, remo te sen sing d rough t mon ito ring mode l can be is sen sitive to ba re so il o r ra re vege ta tion cond ition, bu t no t ve ry effec tive to den se vege ta tion cond ition. d ivided to th ree ca tego rie s: V isib le2N IR mode l, the r2 m a l mode l and m ic rowave mode l. Con side ring so il re2 B ecau se so il mo istu re affec ts vege ta tion evapo ra tion grea tly, it is effec tive to u se the ra tio of app a ren t evap 2 flec tivity dep re ssing w ith the inc rea se of so il mo istu re, V isib le 2N IR mode l ca lcu la te s ind irec tly th rough vege ta2 o ra tion to la ten t evapo ra tion to de sc ribe so il mo istu re Rece ived da te: 2009 207 202; Rev ised da te: 2009 211 210 ( ) Foun da t ion item : Fund s of A gricu ltu re Scien tific and Techno logica l A ch ievem en ts Tran sfo rm a tion 05 EFN2 l6800404 ; Sp ec ial Fund s of Yangtze ( )R ive r Sc ien tific R e sea rch In stitu te YJJ0910 / KJ02 ( ) C orrespon d in g a u thor: L I Zhe 1980 2, m ale, enginee r, Ph. D. , graduated from D ep a rtm en t of Ea rth and Sp ace Sc iences, Pek ing U n ive rsity, Ch i2 na, in 2009. H is m a in research ac tivitie s cover regiona l eco2environm en t mon ito ring and eva lua tion, flood and d rough t d isa s2 ( ) ( ) te r mon ito ring models, d igita l wa te rshed. Te l. 027 282926550 E2m a illizhe@m a il. c rsri. cn ) (( )2 la r sp ec trum , MOD TARN mode ra te re so lu tion tran sunde r den se vege ta tion cond ition. J ack son 1981 ( ) ) ( ) b rough t ou t C rop W a te r Stre ss Index CW S I. Tian m ission, LOW TRAN low re so lu tion tran sm ission , ( ( ) A TCOR a sp a tia lly2adap tive fa st a tmo sp he ric co rrec2 1991 found the re la tion sh ip be tween so il mo istu re ) and CW S I u sing land su rface temp e ra tu re, wa te r evap 2 tion. 6 S a tmo sp he ric co rrec tion mode l is u sed in th is p ap e r. o ra tion and m e teo ro logy da ta. CW S I ha s c lea r p hysica l m ean ing, bu t it requ ire s ground ob se rva tion da ta and The gray num be r of remo te sen sing im age is con2 m e teo ro logy da ta, wh ich dec rea se s its ca lcu la tion p re2 ve rted to im age b righ tne ss wh ich ha s p hysica l m ean ing ( ) c ision. M ic rowave mode l re lie s on the re la tion sh ip a2 Chande r and M a rkham , 2003 . Its fo rm u la tion is mong so il p e rm ittivity, em issivity and so il mo istu re. L= g a in ×DN + offse t = λ M ic rowave remo te sen sing can ob se rve ea rth su rface a ll () L- Lm ax m in )( ()+ L, 1 DN - DN m inm in ()DN- DNday and a ll n igh t, bu t it co st too m uch, and no t ve ry m axm in sen sitive to so il mo istu re unde r den se vege ta tion cond i2 W he re: Lrep re sen ts rad ia tion b righ tne ss of each λ - 2 - 1 - 1 tion. So the re ra ise s a que stion whe the r we can ca rry ( ) μchanne l un it in W ?m ?m ? sr, and offse t and ou t a new d rough t mon ito ring m e thod wh ich is succe ss2 ga in a re regre ssion p a ram e te rs b rough t by im age - 2 - 1 - 1 fu l unde r comp lex unde rlying su rface cond ition. ( ) μheade r un it in W ?m ?m ? sr, DN is co r2 () L and su rface temp e ra tu re L ST, vege ta tion index rec ted gray num be r, and DNa re m axim umDN m in m ax () V Iand a lbedo a re key fac to rs to mon ito r so il mo is2 and m in im um gray num be r re sp ec tive ly, and L Lm inm ax tu re, and the ir comp rehen sive app lica tion can exac tly a re co rre spond ing rad ia tion b righ tne ss of m axim um and - 2 - 1 ( μm in im um gray num be rs un it in W ?m ? m ? reflec t su rface d rough t and wa te r cond ition. Gowa rd - 1 ) ( ) sr. and Hop e 1989 found ou t tha t L ST2NDV I sp ec tra l ( ) sp ace showed a triangle d istribu tion. P rice 1990 be2 Top a tmo sp he ric reflec tivity m ay be exp re ssed a s lieved L ST2NDV I sp ec tra l sp ace d isp layed a trap ezium fo llow s: 2 d istribu tion, then u sed a m a them a tica l fo rm u la tion to π?L?d λ ρ(θθ)( )= . 2 ,, 0TOA s v v de sc ribe it, and app lied it to ana lyse land su rface va2 (θ) E?co s λ s θθW he re: , , and 00s s v v( ) ( ) po riza tion. Zhan 2005 , A buduwa sit 2004 found rep re sen t inc iden t zen ith, in2 ou t tha t N IR 2R ed channe l of ETM + im age s fo rm ed a c iden t azim u th, ob se rva tion zen ith and ob se rva tion azi2 sp ec tra l sp ace wh ich wa s sim ila r to NDV I2L ST sp ec tra l ρ(θθ) m u th re sp ec tive ly, ,, is app a ren t reflec tiv20TOA s v v ( )( ) sp ace, so p e rp end icu la r d rough t index PD I wa s ity a lso ca lled top a tmo sp he ric sp ec tra l reflec tivity b rough t ou t ba sed on N IR 2R ed sp ec tra l sp ace and in2 wh ich is ob ta ined afte r rad ia tion co rrec tion; d deno te s situ so il mo istu re ob se rva tion from N ingxia H u i2na tion sun2ea rth ch ronom e te r un it d istance; Erep re sen ts ave r2 λ A u tonomou s R egion wh ich wa s succe ssfu lly u sed to va l2 λ( age top a tmo sp he ric rad iance in channe lun it in W ? - 2 - 1 ida te PD I mode l. N ingxia H u i2na tion A u tonomou s re2 μ) θ m ?m ; ind ica te s sun zen ith.s gion is loca ted in No rth Ch ina, wh ich ha s lowe r vege ta2 Con side ring a tmo sp he ric ae ro so l m u lti2sca tte r tion cove rage. Can PD I a lso be succe ssfu l unde r com 2 ne ighbo rhood effec t and vapo r ab so rb influence and p lex unde rlying su rface cond ition? Mo tiva ted by th is the re la tion2 p henom enon of an iso trop ic ea rth su rface, que stion, the au tho rs ca rry ou t deep re sea rch. sh ip be tween su rface reflec tivity and app a ren t reflec tiv2 ( ity m ay be exp re ssed a s fo llow s V e rmo te, e t a l. , ) 1997 : 2 M ETHODOLO GY ρ (θ θ ) θ ) ρ (θθ ) (θ,{,, + = t 0 , , 0 TOA v a s v v g ss v v 2. 1 6 S A tm o spher ic C orrec t ion (θ) T s τμ -/v ρρ (θ ) ( )[e + < > t ] } , 3 t d v t ρand 1 - < > S In o rde r to reduce a tmo sp he re in te rfe rence t ρ(θθ) W he re: , , is app a ren t reflec tivity, timp rove re trieva l p rec ision of b iop hysica l p a ram e te r, 0 TOA s v v g (θθ) a tmo sp he ric co rrec tion to remo te sen sing im age is re2 , rep re sen ts a tmo sp he ric ae ro so l m u lti2sca tte rs v (θθ) qu ired. Common ly a tmo sp he ric co rrec tion mode ls in2 ρ and vapo r2ab so rp tion fac to rs; , , 0show s as v v ( c lude 6 S second sim u la tion of sa te llite signa l in the so2 length rad ia tion wh ich is ca rried ou t by R ayle igh sca tte r θand ae ro so l sca tte r; , 0rep re sen t ob se rva tion zen ith v v θ) (and ob se rva tion azim u th re sp ec tive ly; T exp re sse s s ρto ta l tran sm ission of down rad ia tion; <> is environ2 t m en t ne ighbo rhood effec t to su rface p ixe l reflec tivity τμ- / vρ(θ) , e and trep re sen t em issivity of d irec t ra2 t d v d ia tion and up rad ia tion re sp ec tive ly, S is sp he rica l a l2 bedo. 2. 2 N IR2Red Spec tra l Spa ce ( Th is p ap e r u se s the R ed channe l band 3: 630 - ) ( ) 690 nm and N IR channe l band 4: 775 - 900 nm F ig. 1 N IR2Red Spec tra l Spa ce reflec tivitie s of ETM + im age s to con struc t N IR 2R ed ind ica te s R ed channe l reflec tivity; M and I rep re sen t sp ec tra l sp ace. A fte r a tmo sp he ric co rrec tion, R ed coeffic ien ts re sp ec tive ly. channe l and N IR channe l reflec tivitie s of ETM + a re In F ig. 2 , line L wh ich is ve rtica l to B 2C so il line u sed to con struc t a triangle sca tte r m ap , wh ich is sim i2 p a sse s coo rd ina te o rigin, and it can be exp re ssed a s la r to NDV I2L ST sp ec tra l sp ace. Th is triangle sca tte r fo llow s: m ap can no t on ly exp re ss vege ta tion cove rage, bu t a lso ( )×R 5 R= - M . redn ir de sc ribe su rface d rough t. In N IR 2R ed sp ec tra l sp ace, the ve rtica l d istance ( ) R icha rd son and W iegand 1977 bu ilt N IR 2R ed from each po in t to L ine L de sc ribe s su rface d rough t. sp ec tra l sp ace w ith M SS gray va lue s of red channe l and So il mo istu re inc rea se s w ith the ve rtica l d istance de2 N IR channe l, then u sed ve rtica l d istance be tween each c rea se s, and vice ve rsa. To a b lackbody, its ve rtica l po in t and so il line to de sc ribe vege ta tion cove rage, so d istance is ze ro, wh ich m ean s it is loca ted a t coo rd i2 ( ) p e rp end icu la r vege ta tion index PV Iwa s fo rm ed. na te o rigin. O the r non2b lackbody d isp lays h ighe r we t2 F ig. 1 show s the re la tion sh ip be tween N IR 2R ed sp ec tra l ne ss when app roach ing coo rd ina te o rigin. In gene ra l, sp ace and vege ta tion cove rage. L ine A 2E2D ind ica te s sp ace nea r line L rep re sen ts wa te r o r we t ob jec t, and () the vege ta tion cove rage change s from a ll cove r A , to sp ace fa r from line L is d rough t a rea. So the ve rtica l ( ) () p a rtia l cove r E, to ba re so il D . F ig. 1 a lso show s d istance EF from each po in t to L ine L in N IR 2R ed the re la tion sh ip be tween N IR 2R ed sp ec tra l sp ace and sp ec tra l sp ace can be u sed to de sc ribe regiona l so il so il mo istu re. L ine B 2D 2C show s tha t so il we t change s mo istu re, so a remo te sen sing d rough t mon ito ring mod2 () () ( ) from fu ll we t B , to p a rtia l we t D , to d ry C . L ine e l ba sed on N IR 2R ed sp ec tra l sp ace can be given by B 2D 2C is ve rtica l to PV I, and p a ra lle l to so il line. To the fo llow ing fo rm u la: sum up , su rface sp ec tra l fea tu re ha s c lo se re la tion sh ip w ith land cove r and so il mo istu re, and it can be u sed to ind ica te regiona l d rough t. N IR 2R ed sp ec tra l sp ace u se s N IR and red channe l reflec tivity to mon ito r d rough t, wh ich is effec tive and ea sy to p e rfo rm. 2. 3 Rem o te Sen s in g D rough t M ode l in g Ba sed on N IR2Red Spec tra l Spa ce N IR 2R ed sp ec tra l sp ace con struc ted by ETM + re2 mo te sen sing im age s d isp lays a triangle shap e. B 2C is so il line, and so il mo istu re dec rea se s from B to C. 2 so il line can be exM a them a tica l fo rm u la tion of B 2C p re ssed a s fo llow s: ×R ( )4 R= M + I, F ig. 2 Pr in c ip le of Perpen d icu la r D rough t In dex redn ir W he re: Rrep re sen ts N IR channe l reflec tivity; R n irred 1 bod ie s, ba re so il, and o rde red m any ETM + m u lti2 ( ) ( )PD I = R6 + M R . n ir red 2 sp ec tra l remo te sen sing im age s wh ich we re acqu ired M + 1 du ring the sam e p e riod. In the th ird channe l of Zhang2 he Irriga tion System fou r sim u ltaneou s ob se rva tion s 3 M OD EL A PPL ICA T IO N we re dep loyed, of wh ich fou r po sition ing ob se rva tion po in ts we re se lec ted. The re a re cen tra l ob se rva tion 3. 1 Re sea rch A rea ( ) () C1 2C4 , no rthwe st ob se rva tion NW 1 2NW 4 , no rth2 Zhanghe Irriga tion System is loca ted a t the cen te r ( ) ea st ob se rva tion N E1 2N E4 and sou thea st ob se rva2 of H ube i P rovince, Ch ina. Its la titude range s from 30 ? ( ) 00 ′to 31 ?42 ′N , and its longitude range s from 111 ?28 ′tion SE1 2SE4 . Each ob se rva tion po in t dep loyed 4 to 111 5?3 ′E. It tilts from h igh no rthwe st towa rd s low samp le s occup ying 400 squa re m e te rs. Eve ry samp le sou thea st. F rom no rth to sou th, it is abou t 115 km , con ta in s so il con ten t from 2 to 5 cm dep th. So il mo is2 and from ea st to we st, it is abou t 60 km. Its to ta l a rea tu re wa s ca lcu la ted u sing CNC2503DR neu tron mo is2 is app roxim a te ly 5 543. 93 squa re k ilom e te rs. The re2 tu re m e te r. A fte r ca lib ra tion, the we igh t of a lum inum gion is c lay and loam so ils, stagge red in the fram ewo rk box wa s conve rted to the we igh t of wa te r con ten t, of the who le a rea, in wh ich the d istribu tion of c lay is wh ich m ean s un it so il we igh t d ivided by the wa te r it abou t 57 % of the to ta l a rea, loam abou t 43 %. The con ta ined. en tire irriga ted a rea is d ivided in to two p a rts, h ills and Th is p ap e r u se s in2situ ob se rva tion da ta of 2 - 5 p la in s. The h illy a rea is abou t 4 658. 64 squa re k ilo2 cm dep th a t NW 1 , NW 2 , NW 3 and NW 4 on Sep tem 2 m e te rs, accoun ting fo r 84 % of the to ta l a rea, the p la in be r 29 , 2006 , wh ich is sim u ltaneou s w ith sa te llite a rea is abou t 855. 29 squa re k ilom e te rs, accoun ting fo r tran sit. 16 % of the to ta l a rea. The to ta l a rea of cu ltiva ted land 3. 3 Re su lt an d Ana ly s is is 244. 75 m illion ac re s, of wh ich 219. 69 m illion ac re s A cco rd ing to ba re land sp ec tra l reflec tivity info r2 a re p addy fie ld, 25. 06 m illion ac re s a re d ry land. The ( ) m a tion extrac ted by so il line F ig. 3 , and its equa2 impo rtan t gra in2p roduc ing m a in c rop is rice, it is an tion can be exp re ssed a s fo llow s: ba se in H ube i P rovince. = 1. 332 4 ×R - 0. 120 5 ,( )R 7 red n ir is a sub2trop ica l con ti2 Zhanghe Irriga tion System M = 1. 332 4 and I = - 0. 120 5 from above W e extrac t mode ra tion, abundan t nen ta l c lim a te zone, c lim a te equa tion, PD I can be exp re ssed a s fo llow s: ra infa ll, long fro st2free p e riod fo r agricu ltu ra l p roduc2 ( ) PD I = 0. 600 3 ×R+ 1. 332 4 ×R .n ir red tion. Its annua l ave rage temp e ra tu re is 15. 6 - 16. 1 ?. ( )8 The annua l fro st2free p e riod is 246 - 270 days. It ha s abundan t ra infa ll, wh ich change s in la rge, uneven d is2 tribu tion of eve ry yea r. M u lti2yea r ave rage p rec ip ita tion is 968. 5 mm. R a infa ll m a in ly concen tra te from Ap ril to O c tobe r, accoun ting fo r 81. 7 p e rcen t of the who le yea r, in te r2annua l d iffe rence is up to 2. 2 tim e s. The d istribu tion of ra infa ll in the region is unba lance a s F ig. 3 So il l in e ex tra c t ion an d de term in in g of Sou th is b igge r than No rth, and W e st is b igge r than slope pa ram e ter s Ea st. ETM + da ta ha s a h ighe r re so lu tion, so the coo rd i2 3. 2 F ie ld O b serva t ion na te s of ob se rva tion po in ts can be found on geom e tri2 F rom Sep tem be r to O c tobe r in 2006 , the re sea rch2 ca lly co rrec ted ETM im age s. Fo r p u rpo se s of comp a ri2 e rs from Yangtze R ive r Sc ien tific R e sea rch In stitu te ( ) son, fo rm u la 10 is u sed to no rm a lize remo te sen sing ( ) () YR SR I, W uhan U n ive rsity W HU and Ch ine se U 2 d rough t ind ice s and m ea su red so il mo istu re. ( )n ive rsity of Hong Kong CUH K conduc ted a in tegra ted DA TA - DA TA sa te llite 2 land sim u ltaneou s exp e rim en t, co llec ted m in )( DA TA=9 , n DA TA- DA TA m ax m insp ec tra l cha rac te ristic cu rve s of vege ta tion, wa te r W he re: DA TArep re sen ts no rm a lized va lue, DA TA ism ea su red d rough t indexe s. n o rigina l va lue, DA TAand DA TAa re the m axim um m ax m in and m in im um va lue s of a ll samp le re sp ec tive ly. 4 CO NCL US IO N A s it show s in F ig. 4 , 2 - 5 cm in2situ fie ld ob2 Pe rp end icu la r d rough t index m ay reflec t the land se rva tion s and PD I ca lcu la tion s d isp lays ba sica lly the cove r, wa te r2hea t com b ina tion and change s w ith a mo re sam e trend and have a coeffic ien t of 0. 76. In PD I, sp ec ific b iop hysica l m ean ing. It ha s m any advan tage s th is mode l is a comp rehen sive de sc rip to r of bo th so il such a s simp lic ity, p rac tica lity, ea sy2to2u se, ve rsa tility and vege ta tion, no t e rad ica ting vege ta tion in te rfe rence. and un ive rsa lity. In p rac tice, on ly N IR 2R ed reflec2 A lso, CNC 2503DR neu tron mo istu re m e te r is ea sily af2 tance sp ec tra sp ace can be effec tive ly ca rried ou t on fec ted by su rface w ind, c lim a te and hum an ac tivitie s, d rough t mon ito ring in m idd le and lowe r Yangtze R ive r so its re la tion sh ip is no t so good. B a sin. Howeve r, the re a re still two p rob lem s to need fu r2 the r re sea rch. F irstly, a lthough 2 - 5 cm in2situ ob se r2 va tion and PD I ca lcu la tion d isp lay ba sica lly the sam e trend and have a coeffic ien t of 0. 76 , the re la tion sh ip is no t so good, m aybe the rea son lie s in tha t PD I is a comp rehen sive de sc rip to r of bo th so il and vege ta tion, and doe sn’t e rad ica te vege ta tion in te rfe rence. Second2 F ig. 4 C om pa r ison of PD I ca lcu la t ion an d ly, PD I is ba sed on the m a them a tica l exp re ssion of a in 2s itu ob serva t ion fixed so il line, and th is is on ly an idea l a ssump tion. In The ra te of c rop and so il ab so rp tion fie ld cap ac ity, fac t, the shap e of so il line ha s c lo se re la tion sh ip w ith w ilting c rop coeffic ien t and effec tive so il mo istu re con2 so il typ e s and fe rtiliza tion cond ition s. Stric tly sp eak2 ten t a re c lo se ly re la ted to regiona l d rough t cond ition s. ing, so il line is no t nece ssa rily to be a line, wh ich m ay In remo te sen sing d rough t mon ito ring, a comp rehen sive lead to e rro rs. Fu tu re wo rk shou ld focu s on effec tive ly d rough t index com b ined so il mo istu re and wa te r ba l2 sep a ra ting vege ta tion and so il info rm a tion from N IR 2 ance d isp lays a mo re comp rehen sive reflec tion of R ed fea tu re sp ace and imp roving m a them a tica l exp re s2 d rough t situa tion. In2situ m ea su ring d rough t ind ica to r sion equa tion. is p icked up to va lida te PD I. Its m a them a tica l fo rm u la2 REFERENC ES: tion is [ 1 ] WA STON K, ROW EN L C, O FF IELD T W. App lica tion W - W p( )K = 1 - 10 , of the rm a l mode ling in geo logic in te rp re ta tion of IR im age s W - W h p [ J ]. R emo te Sen sing Environm en t, 1971 , 3: 2017 - ( ) W he re: K is d rough t index, W is wa te r mo istu re % , 2041. W is the upp e r lim it tha t so il can m a in ta in, wh ich is h PR ICE J C. O n the ana lysis of the rm a l infra red im age ry: [ 2 ] ( ) a lso ca lled fie ld cap ac ity % , W is so il w ilting coef2 p the lim ited u tility of app a ren t the rm a l ine rtia [ J ]. R e2 ( ) fic ien t % . mo te Sen sing of Environm en t, 1985 , 18: 59 - 93. F ig. 5 show s tha t PD I is con sisten t w ith fie ld [ 3 ] KAHL E A B. A simp le the rm a l mode l of the ea rth’s su r2 face fo r geo logic m app ing by remo te sen sing [ J ]. Jou rna l of Geop hysica l R e sea rch, 1977 , 82: 1673 - 1680. [ 4 ] EN GLAND A W. R ad iob righ tne ss of d iu rna lly hea ted freezing so il [ J ]. IEEE Tran sac tion s on Geo sc ience and ( ) R emo te Sen sing, 1990, 28 3: 464 - 476. [ 5 ] EN GLAND A W , GALAN TOW ICZ J F, SCHR ETTER M S. The rad io b righ tne ss the rm a l ine rtia m ea su re of so il F ig. 5 Re la t ion sh ip be tween PD I an d in 2s itu mo istu re [ J ]. IEEE Tran sac tion s on Geo sc ience and R e2 m ea sured drough t in d ica tor K ( ) mo te Sen sing, 1992, 30 1: 132 - 139. [ 6 ] ZHAN G R en2hua. The rm a l ine rtia mode l and its app lica2 on Geo sc ience and R emo te Sen sing, 1990 , 28: 940 - ( ) tion [ J ]. Ch ine se Sc ience B u lle tin, 1991 , 948. 12: 924 - ( [ 15 ] R IDD L. Exp lo ring a V 2I2S vege ta tion2imp e rviou s su r2 927. 2liang. R emo te sen sing mon ito ring fo r so il T IAN Guo) [ 7 ] face2so ilmode l fo r u rban eco system ana lysis th rough re2 mo istu re [ J ]. R emo te Sen sing of Environm en t, 1991 , 6 mo te sen sing: Comp a ra tive ana tom y fo r c itied [ J ]. In te r2 ( ) na tiona l Jou rna l of R emo te Sen sing, 1995, 16: 2165 - 2 : 89 - 99. [ 8 ] JACKSON R D , ID SO S B. Canop y temp e ra tu re a s a c rop 2185. [ 16 ] G ILL ES S. A ve rifica tion of the triangle m e thod fo r ob ta i2 wa te r stre ss ind ica to r [ J ]. W a te r R e sou rce s R e sea rch, 1981 , 17: 133 - 138. n ing su rface so il wa te r con ten t and ene rgy fluxe s from re2 mo te m ea su rem en ts of the no rm a lized d iffe rence vege ta tion [ 9 ] W IGN ERON J P, CALV ET J C, PELLAR IN T, et a l. () index NDV Iand su rface rad ian t temp e ra tu re [ J ]. In2 R e trieving nea r2su rface so il mo istu re from m ic rowave ra2 ( ) te rna tiona l Jou rna l of R emo te Sen sing, 1997 , 18 5 : d iom e tric ob se rva tion s: Cu rren t sta tu s and fu tu re p lan s 3145 - 3166. [ J ]. R emo te Sen sing of Environm en t, 2003, 85: 489 - 2juan, WAN G Peng2xin, WAN G J in2d i, et a l. HAN L i506. [ 17 ] HU ISMAN J A , HUBBARD S S, R EDMAN J D , et a l. R e sea rch on V I2L ST sp ec tra l sp ace [ J ]. Sc ience in Ch ina [ 10 ] ( ) M ea su ring so il wa te r con ten t w ith ground p ene tra ting ra2 Se rie s D , 2005, 35 4: 371 - 377. GHULAM A , Q IN Q , WAN G L , da r: A review [ J ]. V ado se Zone Jou rna l, 2003, 2: 476 - [ 18 ] et a l. D eve lopm en t of b roadband a lbedo ba sed eco logica l safe ty mon ito ring index 491. [ C ] [ 11 ] CA SH ION J , LA KSHMN I V , BO SCH D , et a l. M ic ro2 ? 2004 IEEE In te rna tiona l Geo sc ience and R emo te ( ) wave remo te sen sing of so il mo istu re: Eva lua tion of the A ncho rage, A la ska, Sen sing Sympo sium IGAR SS , ( ) TRMM m ic rowave im age r TM Isa te llite fo r little rive r Egan Conven tion Cen te r, U SA , Sep tem be r 20 - 24, wa te rshed tifton, geo rgia [ J ]. Jou rna l of H yd ro logy, 2004: 4115 - 4118. 2005 , 37: 242 - 253. 2jing, M a ssayuk i Tam u ra, H ideno ri Taka2 [ 19 ] ZHAO W en [ 12 ] WAN G C, Q I J , MORAN S, et a l. So il mo istu re e stim a2 ha sh i. A tmo sp he ric and sp ec tra l co rrec tion s fo r e stim a ting su rface a lbedo from sa te llite da ta u sing 6S code [ J ]. R etion in a sem ia rid range land u sing ER S22 and TM im age ry 2 [ J ]. R emo te Sen sing of Environm en t, 2004, 90: 178 - mo te Sen sing of Environm en t, 2000 , 76: 202 - 212. 189. [ 20 ] R ICHARD SON A J , W IEGAND C L. D istingu ish ing veg2 [ 13 ] GOWARD S N , HO PE A S. Evapo ra tion from com b ined e ta tion from so il background info rm a tion [ J ]. Pho togram 2 ( ) reflec ted so la r and em itted te rre stria l rad ia tion: P re lim ina2 m e tric Enginee ring and R emo te Sen sing, 1977, 43 12 : 1541 - 1552. ry F IFE re su lts from AVHRR da ta [ J ]. A dvance s in Sp ace R e sea rch, 1989, 9: 239 - 249. [ 14 ] PR ICE J C. U sing sp a tia l con text in sa te llite da ta to infe r ( )Ed ited by L IU Yun2fe i, Y I X in2hua regiona l sca le evapo tran sp ira tion [ J ]. IEEE Tran sac tion s 垂直干旱指数在湖北漳河灌区遥感旱情监测中的应用 11 1 2 穗 李 喆 , 谭德宝 , 秦其明 , 张 ( )1. 长江科学院 空间信息技术应用研究所 , 武汉 430010; 2. 北京大学 地球与空间科学学院 , 北京 100871 摘要 :遥感技术具有覆盖范围广 、空间分辨率高 、重访周期短 、数据获取快捷方便等优点 ,遥感干旱监测已经成为干 ( ) ( ) 旱监测的重要研究方向之一 。詹志明 2005 、阿布 2004 等假设裸露土壤的光谱特征满足线性分布 ,在 N IR 2R ed ( ) 光谱特征空间中提出了垂直干旱指数 PD I,并在我国西北干旱区 2宁夏固原地区进行了模型验证 。为了进一步验 证 PD I在茂密植被覆盖条件下的可行性 ,选择长江中下游流域的典型灌区湖北漳河灌区作为研究区域 ,采用 ETM +数据计算 PD I,与 2,5 cm 平均土壤水分相比较 ,结果表明 :模型观测值变化曲线和实测土壤水分变化趋势保持一致 , 2,5 cm 平均土壤水分相关系数为 0. 76,与实测干旱指标基本吻合 。 PD I模型能够有效地反映地表覆盖和水 热组合 ,简单实用 ,具有通用性和普适性 ,适宜于复杂下垫面条件下的遥感旱情监测 。 关 键 词 : N IR 2R ed光谱特征空间 ; 垂直干旱指数 ; 湖北漳河灌区