Page 66 - 《水资源与水工程学报》2024年第5期
P. 66

!35 "!5 #                         & ' ( ) & * + , -                               Vol.35No.5
               2024 $ 10 %              JournalofWaterResources&WaterEngineering                 Oct.,2024

            DOI:10.11705/j.issn.1672-643X.2024.05.08


                 'xÈZrsvt3d8u÷;<|}ýwXY



                                            mH:, n o, pqr, stu
                                    ( Jaxˆ-Še. AR[ŒÊ§».T, IJ +' 712100)
                 H I: Jahgzf.Fgš, ÿ!ö JahgzfÓÔðxy=ùä%&&+A·#ߦM。• 2001 v 6
                 {—2020 v 5 {{ᬀ$‰ŠäçèîïËA、 ’“、 sGåXh„…{¼E‡ˆ‰Š$@‰ŠE, R†ˆ¸”
                 ˆŽð‹Œçèzfö YV, Î%<YVéJahg{7†­。°±²o: ³YVt{çèzf”‰(CDI) [
                 á¬çèzf”‰(MCI) q{êzfå_¨lûé 75%•Ä;CDI [ᬀ${à!ËA”‰(SPI3) äà!
                 sGW‘”‰(SSI1) {™¨¿‰¢x@ 0.318~0.726 ä 0.173~0.433, µ’ÖFÜÝA´;CDI 1!ªîï
                 Jahg&Vzf^ÊfÓ{&+ÓÔ。%<²o, ˆ¸”ˆ>ð*†ÁJahgçèzfö , A·¨«{l
                 †D½。
                 JKL: zf; ö YV; ˆ¸”ˆ; ¼E‰Š; Jahg
                 MNFOP:S423;TV11   QRSTU:A    QVWP:1672643X(2024)05006211

                         Modelforcomprehensivedroughtmonitoringinnorthwest
                                          Chinabasedonrandom forest

                                   SONGLihui,ZHANGXin,DUChuanyu,ZONGHanxi
                   (SchoolofWaterConservancyandArchitecturalEngineering,NorthwestA&FUniversity,Yangling712100,China)

                 Abstract:NorthwestChinaisafflictedbyfrequentdroughtdisasters,soaccuratemonitoringofdrought
                 conditionsintheregionisofgreatsignificancetoagriculturalproductionandeconomicdevelopment.This
                 articleusesmeteorologicalstationdatafromJune2001toMay2020andmultisourceremotesensingdata
                 setsofprecipitation ,vegetation,soilandotherenvironmentalfactorsasdatasourcestoestablishacom
                 prehensivedroughtmonitoringmodelbasedonrandom forestalgorithm.Then,theapplicabilityofthe
                 modelinthenorthwestChinaisinvestigated.Theresultsshowthattheconsistencyrateofthedroughtlev
                 elsineachmonthofthestudyperiodbetweenthecomprehensivedroughtindex(CDI)outputbythe
                 modelandthemeteorologicalcomprehensivedroughtindex (MCI)isabove75%.Moreover,thecorrela
                 tioncoefficientsbetweenCDIandthestandardizedprecipitationindex(SPI3),CDIandthestandard
                 izedsoilmoistureindex(SSI1)ofthemeteorologicalstationsare0.318-0.726and0.173-0.433re
                 spectively ,bothreachingextremelysignificantlevels.Inaddition,CDIcanaccuratelyreflectthedevel
                 opmentofdroughtconditionsintypicaldroughteventsinthenorthwestregion.Thisstudyshowsthatthe
                 randomforestmethodisapplicabletothecomprehensivedroughtmonitoringinnorthwestChina.
                 Keywords:drought;monitoringmodel;randomforest(RF);multisourcedata;northwestChina

                                                               äáõMNñò, A·ËA=œ、 ¢dºµÃT&Ï
            1 ./01
                                                               ø{€$    [2] , Olzf.Fg&, ³÷Jahg
                 zf.FX+,.FQi4%&•bôÇ#、                          {xy=ùä%&&+           [3] 。„8, ÿ!ö zfð
            ‚ƒefô¢{á¬.F             [1] 。JahghžXh              úàJahgzfö 12A·#ߦM, u]·


                !"#$:20240421; %&#$:20240625
                '()*: !"+,-./0)*(52379026)
                +,-.: ½wl(1998—), H, ;;õø9, u;%<=, %<>?@ABADE。
                /0+,: c °(1968—), H, 56ön9, :;, MN, u;=OP, Þߍfáõ~[AWX、 zfwx[|}˜~™
                        š%<。
   61   62   63   64   65   66   67   68   69   70   71