{"id":2723,"date":"2016-12-06T14:28:39","date_gmt":"2016-12-06T14:28:39","guid":{"rendered":"http:\/\/www.vtei.cz\/?p=2723"},"modified":"2024-07-16T11:52:57","modified_gmt":"2024-07-16T10:52:57","slug":"quantifying-the-retention-characteristics-by-the-means-of-geomorphological-patterns-of-the-basin","status":"publish","type":"post","link":"https:\/\/www.vtei.cz\/en\/2016\/12\/quantifying-the-retention-characteristics-by-the-means-of-geomorphological-patterns-of-the-basin\/","title":{"rendered":"Quantifying the retention characteristics by the means of geomorphological patterns of the basin"},"content":{"rendered":"

<\/i> This article is available in Czech only. For translation or more information on this topic, please contact author.<\/h4>\n

 <\/p>\n

Souhrn<\/h2>\n

V\u00a0p\u0159edlo\u017een\u00e9 studii byly testov\u00e1ny vybran\u00e9 line\u00e1rn\u00ed a\u00a0neline\u00e1rn\u00ed regresn\u00ed modely, kter\u00e9 popisuj\u00ed vz\u00e1jemn\u00fd vztah mezi parametry modelu Bilan, kter\u00e9 jsou stanoveny na z\u00e1klad\u011b meteorologick\u00fdch a\u00a0hydrologick\u00fdch \u0159ad a\u00a0mezi reten\u010dn\u00edmi charakteristikami stanoven\u00fdmi na z\u00e1klad\u011b vybran\u00fdch geomorfologick\u00fdch charakteristik povod\u00ed.<\/p>\n\"img_2478_uprava\"<\/a>\n

Odvozen\u00e9 line\u00e1rn\u011b regresn\u00ed modely umo\u017e\u0148uj\u00ed odhad parametr\u016f modelu Bilan na z\u00e1klad\u011b reten\u010dn\u00edch charakteristik, kter\u00e9 odhaduj\u00ed retenci v\u00a0plo\u0161e povod\u00ed S3, a\u00a0charakteristik, kter\u00e9 odhaduj\u00ed retenci v\u00a0niv\u011b povod\u00ed Smax, St a\u00a0Dd. Uveden\u00fd p\u0159\u00edsp\u011bvek prezentuje prvn\u00ed v\u00fdsledky region\u00e1ln\u00ed anal\u00fdzy zam\u011b\u0159en\u00e9 na kvantifikaci reten\u010dn\u00edch charakteristik anal\u00fdzou geomorfologick\u00fdch charakteristik povod\u00ed.<\/p>\n

\u00davod<\/h2>\n

Retence vody v\u00a0povod\u00ed je jedn\u00edm z\u00a0kl\u00ed\u010dov\u00fdch faktor\u016f, kter\u00e9 utv\u00e1\u0159ej\u00ed v\u00fdslednou hydrologickou bilanci povod\u00ed. Celkovou retenci vody pro dan\u00e9 povod\u00ed tvo\u0159\u00ed \u010dasov\u011b a\u00a0prostorov\u011b prom\u011bnliv\u00e9 z\u00e1soby vody zadr\u017een\u00e9 ve sn\u011bhov\u00e9 pokr\u00fdvce, na povrchu vegeta\u010dn\u00edho pokryvu, na povrchu povod\u00ed, v\u00a0nenasycen\u00e9 a\u00a0nasycen\u00e9 z\u00f3n\u011b.<\/p>\n

Jeden ze z\u00e1kladn\u00edch p\u0159\u00edstup\u016f kvantifikace dynamiky celkov\u00e9 retence vody v\u00a0povod\u00ed je zalo\u017een na aplikaci matematick\u00e9ho modelu hydrologick\u00e9 bilance, kter\u00fd umo\u017e\u0148uje kontinu\u00e1ln\u00ed simulaci hydrologick\u00e9 bilance pro vybran\u00fd \u0159e\u0161en\u00fd \u010dasov\u00fd \u00fasek. Celkov\u00e1 retence je n\u00e1sledn\u011b stanovena na z\u00e1klad\u011b podrobn\u00e9 anal\u00fdzy simulovan\u00fdch tok\u016f a\u00a0z\u00e1sob v\u00fdsledn\u00e9 hydrologick\u00e9 bilance povod\u00ed. V\u00a0podm\u00ednk\u00e1ch \u010cesk\u00e9 republiky je pro pot\u0159eby \u0159e\u0161en\u00ed cel\u00e9 \u0159ady souvisej\u00edc\u00edch in\u017een\u00fdrsk\u00fdch \u00faloh vyu\u017e\u00edv\u00e1n celistv\u00fd model hydrologick\u00e9 bilance Bilan [1, 2]. Model Bilan umo\u017e\u0148uje popsat dynamiku jednotliv\u00fdch z\u00e1sob vody v\u00a0povod\u00ed, kter\u00e1 je vyj\u00e1d\u0159ena jejich \u010dasov\u00fdmi \u0159adami. K\u00a0tomu \u00fa\u010delu jsou pou\u017e\u00edv\u00e1ny vybran\u00e9 \u010dasov\u00e9 \u0159ady meteorologick\u00fdch a\u00a0hydrologick\u00fdch veli\u010din.<\/p>\n

Vedle simulac\u00ed celkov\u00e9 retence kontinu\u00e1ln\u00edm modelem hydrologick\u00e9 bilance existuje cel\u00e1 \u0159ada empirick\u00fdch p\u0159\u00edstup\u016f, kter\u00e9 poskytuj\u00ed odhad retence vody v\u00a0povod\u00ed. Jedn\u00edm z\u00a0nejroz\u0161\u00ed\u0159en\u011bj\u0161\u00edch je empirick\u00fd postup, kter\u00fd vyu\u017e\u00edv\u00e1 \u010d\u00edsla odtokov\u00fdch k\u0159ivek (CN k\u0159ivky) pro stanoven\u00ed odhadu maxim\u00e1ln\u00ed retence povod\u00ed [3].<\/p>\n

D\u00e1le cel\u00e1 \u0159ada studi\u00ed poukazuje na vz\u00e1jemn\u00fd vztah mezi hydrologick\u00fdm syst\u00e9mem povod\u00ed a\u00a0geomorfologick\u00fdmi charakteristikami povod\u00ed [4\u20136]. Tyto vztahy je mo\u017en\u00e9 pou\u017e\u00edt pro regionalizaci hodnot vybran\u00fdch parametr\u016f hydrologick\u00e9ho modelu, kter\u00e1 je zalo\u017eena na vyu\u017eit\u00ed fyzik\u00e1ln\u00edch geomorfologick\u00fdch charakteristik povod\u00ed [4, 5]. V\u00a0p\u0159\u00edpad\u011b existence vz\u00e1jemn\u00e9ho vztahu mezi geomorfologick\u00fdmi charakteristikami povod\u00ed a\u00a0parametry hydrologick\u00e9ho modelu je mo\u017en\u00e9 kvantifikovat vliv nov\u011b navrhovan\u00fdch reten\u010dn\u00edch opat\u0159en\u00ed [7]. Podm\u00ednkou kvantifikace je existence vztahu mezi reten\u010dn\u00edmi opat\u0159en\u00edmi a\u00a0geomorfologick\u00fdmi charakteristikami [4].<\/p>\n

C\u00edlem p\u0159edlo\u017een\u00e9ho p\u0159\u00edsp\u011bvku je popis vz\u00e1jemn\u00e9ho vztahu vybran\u00fdch geomorfologick\u00fdch charakteristik povod\u00ed a\u00a0vybran\u00fdch parametr\u016f hydrologick\u00e9ho modelu Bilan, kter\u00e9 nejv\u00edce ovliv\u0148uj\u00ed celkovou retenci vody v\u00a0povod\u00ed. Na z\u00e1klad\u011b odvozen\u00fdch vztah\u016f bude n\u00e1sledn\u011b mo\u017en\u00e9 kvantifikovat vliv vybran\u00fdch reten\u010dn\u00edch opat\u0159en\u00ed v\u00a0povod\u00ed na celkovou retenci vody v\u00a0povod\u00ed.<\/p>\n

Materi\u00e1l a\u00a0metody<\/h2>\n

Data povod\u00ed LAPV<\/h3>\n

V\u00a0p\u0159edlo\u017een\u00e9 studii byly pou\u017eity dva druhy dat. Prvn\u00ed byl tvo\u0159en m\u011bs\u00ed\u010dn\u00edmi meteorologick\u00fdmi a\u00a0hydrologick\u00fdmi \u0159adami pro povod\u00ed \u0159e\u0161en\u00fdch v\u00a0r\u00e1mci Generelu lokalit pro akumulaci povrchov\u00fdch vod (LAPV). Druh\u00fd soubor dat je tvo\u0159en prostorov\u00fdmi daty, kter\u00e1 umo\u017e\u0148uj\u00ed stanovit vybran\u00e9 charakteristiky povod\u00ed.<\/p>\n

Pro ka\u017ed\u00e9 povod\u00ed byla provedena kalibrace hydrologick\u00e9ho modelu Bilan, jej\u00edm\u017e v\u00fdsledkem bylo stanoven\u00ed hodnot parametr\u016f modelu Bilan pro dan\u00e9 povod\u00ed LAPV. V\u00a0p\u0159edlo\u017een\u00e9 studii byly testov\u00e1ny parametry Spa a\u00a0Grd.<\/p>\n\"maca-1\"<\/a>\n

Obr.\u00a01. V\u00fdsledky line\u00e1rn\u00edho modelu krokov\u00e9 regrese pro parametr Spa; A) shoda mezi Spa stanoven\u00fdm na z\u00e1klad\u011b kalibrace modelu Bilan a\u00a0Spa stanoven\u00fdm geomorfologickou anal\u00fdzou; B) z\u00e1vislost Spa stanoven\u00fdm geomorfologickou anal\u00fdzou na pr\u016fm\u011brn\u00e9m sklonu \u0159\u00ed\u010dn\u00ed s\u00edt\u011b St
\nFig. 1. Results of stepwise linear regression model for Spa parameter; A) scatter plot of Spa estimated using the Bilan model and Spa estimated using geomorphological analysis; B) relationship between Spa estimated using geomorphological analysis and average slope of river network<\/h6>\n

Parametr Spa vyjad\u0159uje velikost neline\u00e1rn\u00edho z\u00e1sobn\u00edku z\u00e1soby p\u016fdn\u00ed vody a\u00a0intercepce v\u00a0povod\u00ed v\u00a0mm, kter\u00fd ur\u010duje \u010d\u00e1st retence vody v\u00a0povod\u00ed. Zv\u00fd\u0161en\u00edm hodnot Spa je zv\u00fd\u0161ena uveden\u00e1 \u010d\u00e1st celkov\u00e9 retence vody v\u00a0povod\u00ed. Parametr Grd je sou\u010dinitel z\u00e1sobnosti akumula\u010dn\u00edch prostor\u016f, kter\u00e9 kontroluj\u00ed z\u00e1kladn\u00ed odtok. Sn\u00ed\u017een\u00edm hodnot Grd doch\u00e1z\u00ed ke zpomalen\u00ed z\u00e1kladn\u00edho odtoku v\u00a0povod\u00ed.<\/p>\n

Anal\u00fdza vz\u00e1jemn\u00e9ho vztahu geomorfologick\u00fdch charakteristik a\u00a0parametr\u016f modelu Bilan byla provedena na souboru vybran\u00fdch 53 povod\u00ed LAPV. Na z\u00e1klad\u011b GIS anal\u00fdz byly stanoveny pro ka\u017ed\u00e9 povod\u00ed n\u00e1sleduj\u00edc\u00ed geomorfologick\u00e9 charakteristiky povod\u00ed: A\u00a0\u2013 plocha povod\u00ed [km2<\/sup>]; CN1, CN2, CN3\u00a0\u2013 \u010d\u00edsla odtokov\u00fdch k\u0159ivek pro t\u0159i p\u0159edchoz\u00ed vl\u00e1hov\u00e9 pom\u011bry; S1, S2, S3\u00a0\u2013 maxim\u00e1ln\u00ed retence povod\u00ed stanoven\u00e1 metodou CN pro t\u0159i typy p\u0159edchoz\u00edch vl\u00e1hov\u00fdch pom\u011br\u016f [mm]; Dd\u00a0\u2013 hustota \u0159\u00ed\u010dn\u00ed s\u00edt\u011b [km\/km2<\/sup>]; L\u00a0\u2013 celkov\u00e1 d\u00e9lka \u0159\u00ed\u010dn\u00ed s\u00edt\u011b [km]; S\u00a0\u2013 pr\u016fm\u011brn\u00fd sklon \u0159\u00ed\u010dn\u00ed s\u00edt\u011b [%]; Sp\u00a0\u2013 sklon povod\u00ed [%], Smax\u00a0\u2013 maxim\u00e1ln\u00ed lok\u00e1ln\u00ed sklon povod\u00ed [%].<\/p>\n

Regresn\u00ed modely<\/h3>\n

Pro stanoven\u00ed vz\u00e1jemn\u00e9ho vztahu mezi vybran\u00fdmi parametry modelu Bilan (Spa a\u00a0Grd) a\u00a0geomorfologick\u00fdmi charakteristikami povod\u00ed (deskriptory) byly pou\u017eity n\u00e1sleduj\u00edc\u00ed regresn\u00ed modely: model v\u00edcerozm\u011brn\u00e9 line\u00e1rn\u00ed regrese, model krokov\u00e9 regrese a\u00a0neline\u00e1rn\u00ed regresn\u00ed model zalo\u017een\u00fd na Gaussov\u011b procesu [8].<\/p>\n

V\u00edcerozm\u011brn\u00fd model line\u00e1rn\u00ed regrese ur\u010duje ka\u017ed\u00fd parametr modelu Bilan pomoc\u00ed line\u00e1rn\u00ed kombinace hodnot vybran\u00fdch geomorfologick\u00fdch charakteristik a\u00a0nezn\u00e1m\u00fdch parametr\u016f, kter\u00e9 byly stanoveny standardn\u00ed metodou nejmen\u0161\u00edch \u010dtverc\u016f [9].<\/p>\n

Model krokov\u00e9 regrese pomoc\u00ed itera\u010dn\u00edho v\u00fdpo\u010dtu identifikuje vhodn\u00fd soubor geomorfologick\u00fdch charakteristik, kter\u00e9 vysv\u011btluj\u00ed testovan\u00e9 parametry modelu Bilan. Pro pot\u0159eby \u0159e\u0161en\u00ed byl pou\u017eit jednoduch\u00fd prunning algoritmus [10], kter\u00fd itera\u010dn\u00ed v\u00fdpo\u010det za\u010d\u00edn\u00e1 s\u00a0modelem s\u00a0nejv\u011bt\u0161\u00edm po\u010dtem deskriptor\u016f. V\u00a0ka\u017ed\u00e9 iteraci je n\u00e1sledn\u011b postupn\u011b odeb\u00edr\u00e1n\/p\u0159id\u00e1v\u00e1n deskriptor s\u00a0nejmen\u0161\u00edm pod\u00edlem na vysv\u011btlen\u00ed celkov\u00e9ho rozptylu na z\u00e1klad\u011b v\u00fdsledk\u016f F\u00ad\u2011testu [9]. Itera\u010dn\u00ed v\u00fdpo\u010det kon\u010d\u00ed nalezen\u00edm vhodn\u00e9ho souboru statisticky v\u00fdznamn\u00fdch deskriptor\u016f. Pro v\u00fdpo\u010det modelem krokov\u00e9 regrese byly pou\u017eity, v\u00a0r\u00e1mci ka\u017ed\u00e9 iterace, modely v\u00edcerozm\u011brn\u00e9 line\u00e1rn\u00ed regrese. V\u00fdsledn\u00fd model krokov\u00e9 regrese je tedy v\u00edcerozm\u011brn\u00fd model line\u00e1rn\u00ed regrese, kter\u00fd p\u0159edpov\u00edd\u00e1 parametry modelu Bilan souborem statisticky v\u00fdznamn\u00fdch deskriptor\u016f geomorfologick\u00fdch charakteristik povod\u00ed.<\/p>\n

Model neline\u00e1rn\u00ed regrese zalo\u017een\u00fd na Gaussov\u011b procesu je pops\u00e1n funkc\u00ed st\u0159edn\u00ed hodnoty a\u00a0kovarian\u010dn\u00ed funkc\u00ed analyzovan\u00e9ho souboru dat. Pro vyj\u00e1d\u0159en\u00ed kovarian\u010dn\u00ed funkce byly pou\u017eity RBF funkce. Identifikace parametr\u016f byla provedena metodou viz [11]. Pro stanoven\u00ed hodnot parametr\u016f Spa a\u00a0Grd byly testov\u00e1ny soubory r\u016fzn\u00fdch kombinac\u00ed dvou a\u00a0t\u0159\u00ed geomorfologick\u00fdch deskriptor\u016f.<\/p>\n

Cel\u00fd v\u00fdpo\u010det byl proveden ve statistick\u00e9m prost\u0159ed\u00ed R, regresn\u00ed modely Gaussova procesu byly po\u010d\u00edt\u00e1ny pomoc\u00ed bal\u00edku kernlab [12].<\/p>\n

V\u00fdsledky<\/h2>\n

Kalibrace modelu Bilan pro povod\u00ed LAPV<\/h3>\n

Pro kalibraci modelu Bilan byla vyu\u017eita data 53 povod\u00ed zpracovan\u00fdch v\u00a0r\u00e1mci Generelu lokalit pro akumulaci povrchov\u00fdch vod (Generel LAPV). Pro stanoven\u00ed hodnot parametr\u016f Spa a\u00a0Grd byl pou\u017eit manu\u00e1ln\u00ed a\u00a0automatick\u00fd kalibra\u010dn\u00ed postup. Kvalita kalibrace byla ohodnocena na z\u00e1klad\u011b porovn\u00e1n\u00ed pozorovan\u00fdch a\u00a0vypo\u010dten\u00fdch hodnot m\u011bs\u00ed\u010dn\u00edch v\u00fd\u0161ek odtoku koeficientem determinace (NS) a\u00a0hodnotami pr\u016fm\u011brn\u00e9 absolutn\u00ed odchylky (MAE).<\/p>\n

Hodnoty koeficientu determinace se nach\u00e1zely v\u00a0intervalu (0,03; 0,75). Pr\u016fm\u011brn\u00e1 hodnota NS byla rovna 0,48, medi\u00e1n 0,51, 50\u2005% povod\u00ed bylo nakalibrov\u00e1no s\u00a0hodnotami NS mezi 0,38 a\u00a00,59, u\u00a025\u2005% povod\u00ed byly v\u00fdsledky kalibrace ohodnoceny hodnotou NS mezi 0,59 a\u00a00,75.<\/p>\n

Hodnoty pr\u016fm\u011brn\u00e9 absolutn\u00ed odchylky se nach\u00e1zely v\u00a0intervalu (1,91; 13,68). Pr\u016fm\u011brn\u00e1 hodnota MAE byla rovna 7,64\u2008mm, medi\u00e1n 6,73\u2008mm, 50\u2005% povod\u00ed bylo nakalibrov\u00e1no s\u00a0hodnotami MAE mezi 5,50\u2008mm a\u00a09,83\u2008mm, u\u00a025\u2005% povod\u00ed byly v\u00fdsledky kalibrace ohodnoceny hodnotou MAE mezi 1,91\u2008mm a\u00a05,50\u2008mm.<\/p>\n

Identifikovan\u00e9 hodnoty parametru Spa se pohybovaly v\u00a0rozp\u011bt\u00ed 23\u2008mm a\u00a0163\u2008mm, pr\u016fm\u011brn\u00e1 hodnota Spa byla rovna 84,13\u2008mm, medi\u00e1n 80,43\u2008mm, 50\u2005% povod\u00ed m\u011blo hodnotu Spa mezi 64,39\u2008mm a\u00a0108,60\u2008mm, u\u00a025\u2005% povod\u00ed byla celkov\u00e1 retence p\u016fdn\u00edho a\u00a0intercep\u010dn\u00edho z\u00e1sobn\u00edku mezi 108,60\u2008mm a\u00a0163\u2008mm, pro 25\u2005% povod\u00ed s\u00a0nejni\u017e\u0161\u00ed hodnotou Spa byla hodnota parametru nalezena mezi 23\u2008mm a\u017e 64\u2008mm.<\/p>\n

Identifikovan\u00e9 hodnoty parametru Grd se nach\u00e1zely v\u00a0rozp\u011bt\u00ed 0,05 a\u00a00,50\u2008mm, pr\u016fm\u011brn\u00e1 hodnota Grd byla rovna 0,52, medi\u00e1n 0,25, 50\u2005% povod\u00ed m\u011blo hodnotu Grd mezi 0,20 a\u00a00,31, u\u00a025\u2005% povod\u00ed byla hodnota sou\u010dinitele z\u00e1sobnosti pro z\u00e1sobn\u00edk podzemn\u00ed vody mezi 0,31 a\u00a00,50, pro 25\u2005% povod\u00ed s\u00a0nejni\u017e\u0161\u00ed hodnotou Grd byla hodnota parametru nalezena mezi 0,05 a\u017e 0,20.<\/p>\n

Krokov\u00e1 regrese<\/h3>\n

Pro stanoven\u00ed parametru Spa byl identifikov\u00e1n n\u00e1sleduj\u00edc\u00ed line\u00e1rn\u00ed model na z\u00e1klad\u011b krokov\u00e9 regrese:
\nSpa = 0,71 S3 + 88,91 Dd\u00a0\u2013 8,89 St + 65,76\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 (1)
\nkde hlavn\u00ed soubor deskriptor\u016f je tvo\u0159en S3 maxim\u00e1ln\u00ed retenc\u00ed pro III. typ p\u0159edchoz\u00edch vl\u00e1hov\u00fdch podm\u00ednek [mm]; Dd je hustota \u0159\u00ed\u010dn\u00ed s\u00edt\u011b povod\u00ed [km\/km2<\/sup>] a\u00a0St pr\u016fm\u011brn\u00fd sklon \u0159\u00ed\u010dn\u00ed s\u00edt\u011b v\u00a0[%]. Cel\u00fd soubor p\u0159edstavuje statisticky v\u00fdznamn\u00e9 vstupn\u00ed veli\u010diny, co\u017e bylo potvrzeno v\u00fdsledky F testu a\u00a0t\u00ad\u2011test\u016f, koeficient determinace line\u00e1rn\u00edho modelu je roven 0,33. V\u00fdslednou z\u00e1vislost ukazuje obr.\u00a01 a<\/em>. Nejt\u011bsn\u011bj\u0161\u00ed line\u00e1rn\u00ed shodu mezi hodnotami Spa a\u00a0pou\u017eit\u00fdmi deskriptory vykazuje vztah mezi sklonem toku a\u00a0Spa (viz\u00a0obr.\u00a01<\/em>\u2008b<\/em>).<\/p>\n\"maca-2\"<\/a>\n

Obr.\u00a02. V\u00fdsledky line\u00e1rn\u00edho modelu krokov\u00e9 regrese pro parametr Grd; A) shoda mezi Grd stanoven\u00fdm na z\u00e1klad\u011b kalibrace modelu Bilan a\u00a0Grd stanoven\u00fdm geomorfologickou anal\u00fdzou; B) z\u00e1vislost Grd stanoven\u00fdm geomorfologickou anal\u00fdzou na pr\u016fm\u011brn\u00e9m sklonu \u0159\u00ed\u010dn\u00ed s\u00edt\u011b St
\nFig. 2. Results of stepwise linear regression model for Grd parameter; A) scatter plot of Grd estimated using the Bilan model and Grd estimated using geomorphological analysis; B) relationship between Grd estimated using geomorphological analysis and average slope of river network<\/h6>\n

Pro parametr Grd byl krokovou regres\u00ed stanoven n\u00e1sleduj\u00edc\u00ed line\u00e1rn\u00ed model:
\nGrd = 0,02 St\u00a0\u2013 0,003 Smax + 0,23\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 (2)
\nkde soubor deskriptor\u016f je tvo\u0159en St pr\u016fm\u011brn\u00fdm sklonem \u0159\u00ed\u010dn\u00ed s\u00edt\u011b v\u00a0[%], Smax maxim\u00e1ln\u00edm lok\u00e1ln\u00edm sklonem v\u00a0povod\u00ed v\u00a0[%]. Op\u011bt v\u00fdsledn\u00fd model obsahuje statisticky v\u00fdznamn\u00e9 vysv\u011btluj\u00edc\u00ed geomorfologick\u00e9 veli\u010diny podle F testu a\u00a0t\u00ad\u2011test\u016f, koeficient determinace je roven 0,17 (viz\u00a0obr.\u00a02<\/em>\u2008a<\/em>). Nejt\u011bsn\u011bj\u0161\u00ed shodu deskriptor\u016f a\u00a0hodnot Grd op\u011bt vykazoval vztah mezi Grd a\u00a0sklonem \u0159\u00ed\u010dn\u00edho syst\u00e9mu St (viz\u00a0obr.\u00a02<\/em>\u2008b<\/em>).<\/p>\n

Jak je patrn\u00e9 ze z\u00e1vislost\u00ed na obr.\u00a01<\/em>\u2008b<\/em> a\u00a02<\/em>\u2008b<\/em> p\u0159i implementaci revitaliza\u010dn\u00edch opat\u0159en\u00ed, kter\u00e9 vedou ke sn\u00ed\u017een\u00ed pod\u00e9ln\u00e9ho sklonu tok\u016f v\u00a0povod\u00ed, bude zv\u00fd\u0161ena celkov\u00e1 retence v\u00a0povod\u00ed. Rovnice 1 a\u00a02 umo\u017e\u0148uj\u00ed kvantifikovat retenci pomoc\u00ed odhadu hodnot parametr\u016f Spa a\u00a0Grd a\u00a0n\u00e1sledn\u00e9m vyu\u017eit\u00ed p\u0159i simulaci hydrologick\u00e9 bilance modelem Bilan.<\/p>\n

Neline\u00e1rn\u00ed regrese\u00a0\u2013 Gaussovsk\u00fd proces<\/h3>\n

Na z\u00e1klad\u011b testov\u00e1n\u00ed r\u016fzn\u00fdch neline\u00e1rn\u00edch regresn\u00edch model\u016f byly pro oba parametry vybr\u00e1ny n\u00e1sleduj\u00edc\u00ed modely. Parametr Spa byl odhadov\u00e1n pomoc\u00ed St pr\u016fm\u011brn\u00e9ho sklonu \u0159\u00ed\u010dn\u00ed s\u00edt\u011b v\u00a0[%] a\u00a0hodnot odtokov\u00fdch k\u0159ivek CN2 pro druh\u00fd typ p\u0159edchoz\u00edch vl\u00e1hov\u00fdch pom\u011br\u016f. Parametr Grd byl odhadov\u00e1n na z\u00e1klad\u011b celkov\u00e9 d\u00e9lky \u0159\u00ed\u010dn\u00ed s\u00edt\u011b v\u00a0povod\u00ed Lt [km] a\u00a0hodnot odtokov\u00fdch k\u0159ivek CN3 pro t\u0159et\u00ed typ p\u0159edchoz\u00edch vl\u00e1hov\u00fdch pom\u011br\u016f.<\/p>\n\"maca-3\"<\/a>\n

Obr.\u00a03. Kalibrace a\u00a0validace neline\u00e1rn\u00edho regresn\u00edho modelu, \u010dern\u011b jsou zn\u00e1zorn\u011bny v\u00fdsledky kalibrace, \u010derven\u011b jsou zobrazeny v\u00fdsledky validace
\nFig. 3. Calibration and validation results of nonlinear regression model, black dots show the calibration dataset and the red dots show the validation dataset<\/h6>\n

Valida\u010dn\u00ed postup neline\u00e1rn\u00edch regresn\u00edch model\u016f, zalo\u017een\u00fdch na Gaussovsk\u00e9m procesu, se skl\u00e1dal z\u00a0kalibrace regresn\u00edch model\u016f na kalibra\u010dn\u00edch souborech dat, validace byla provedena pro osm n\u00e1hodn\u011b vybran\u00fdch povod\u00ed, kter\u00e1 nebyla zahrnuta do kalibra\u010dn\u00edho souboru. Valida\u010dn\u00ed v\u00fdsledky model\u016f obou parametr\u016f ukazuj\u00ed grafy na obr.\u00a03<\/em>.<\/p>\n

Kvalita kalibrace regresn\u00edho modelu pro Spa byla kvantifikov\u00e1na line\u00e1rn\u00edm modelem mezi hodnotami parametru SpaBILAN<\/sub> stanoven\u00fdmi na z\u00e1klad\u011b meteorologick\u00fdch a\u00a0hydrologick\u00fdch \u010dasov\u00fdch \u0159ad a\u00a0hodnotami SpaGAUSSPR<\/sub> stanoven\u00fdmi na z\u00e1klad\u011b geomorfologick\u00fdch charakteristik. Koeficient determinace pro kalibra\u010dn\u00ed soubor byl roven 0,81, pro valida\u010dn\u00ed soubor 0,82. MAE byla rovna 11\u2008mm pro kalibra\u010dn\u00ed soubor a\u00a025\u2008mm pro valida\u010dn\u00ed soubor.<\/p>\n

Kvalita regresn\u00edho modelu pro parametr Grd byla op\u011bt ohodnocena line\u00e1rn\u00edm regresn\u00edm modelem mezi GrdBILAN<\/sub> stanoven\u00fdm na z\u00e1klad\u011b meteorologick\u00fdch a\u00a0hydrologick\u00fdch \u010dasov\u00fdch \u0159ad a\u00a0hodnotami GrdGAUSSPR<\/sub> stanoven\u00fdmi na z\u00e1klad\u011b geomorfologick\u00fdch charakteristik. Koeficient determinace pro kalibra\u010dn\u00ed soubor byl roven 0,74, pro valida\u010dn\u00ed soubor 0,72. MAE byla rovna 0,04 pro kalibra\u010dn\u00ed soubor a\u00a00,05 pro valida\u010dn\u00ed soubor.<\/p>\n

V\u00fdsledn\u00fd vztah mezi parametry odvozen\u00fdmi na z\u00e1klad\u011b pou\u017eit\u00fdch geomorfologick\u00fdch charakteristik ukazuje obr.\u00a04<\/em>.<\/p>\n

Diskuse<\/h2>\n

Podle v\u00fdsledk\u016f kalibrace byl \u00fasp\u011b\u0161n\u011bj\u0161\u00ed neline\u00e1rn\u00ed regresn\u00ed model ne\u017e line\u00e1rn\u00ed regresn\u00ed model. P\u0159i hled\u00e1n\u00ed vhodn\u00e9ho neline\u00e1rn\u00ed regresn\u00edho modelu byly zkou\u0161eny r\u016fzn\u00e9 soubory deskriptor\u016f. Uveden\u00e9 modely byly \u00fasp\u011b\u0161n\u011bj\u0161\u00ed ne\u017e zm\u00edn\u011bn\u00e9 line\u00e1rn\u00ed modely. Podobn\u00e9 v\u00fdsledky byly dosa\u017eeny v\u00a0pr\u00e1ci [5].<\/p>\n\"maca-4\"<\/a>\n

Obr.\u00a04. Vztah mezi Spa, pr\u016fm\u011brn\u00fdm sklonem \u0159\u00ed\u010dn\u00ed s\u00edt\u011b a\u00a0CN2 \u010d\u00edslem odtokov\u00fdch k\u0159ivek (lev\u00e1 \u010d\u00e1st obr\u00e1zku); vztah mezi Grd d\u00e9lkou \u0159\u00ed\u010dn\u00ed s\u00edt\u011b a\u00a0\u010d\u00edslem odtokov\u00fdch k\u0159ivek CN3 (prav\u00e1 \u010d\u00e1st obr\u00e1zku)
\nFig. 4. The relationship between Spa and average slope of river network St and curve number CN2 (left part of the figure); the relationship between Grd and length of river network and curve number CN3 (right part of the figure)<\/h6>\n

Skladba deskriptor\u016f a\u00a0odvozen\u00e9 vztahy jsou logick\u00e9, nap\u0159.\u00a0zv\u00fd\u0161en\u00ed z\u00e1sob podzemn\u00ed vody m\u016f\u017ee b\u00fdt dosa\u017eeno revitaliza\u010dn\u00edm opat\u0159en\u00edm, jeho\u017e z\u00e1m\u011brem je zv\u011bt\u0161en\u00ed celkov\u00e9 d\u00e9lky tok\u016f, v\u00a0r\u00e1mci nich\u017e je p\u0159edpokl\u00e1d\u00e1no zv\u00fd\u0161en\u00ed infiltrace v\u00a0nivn\u00edm syst\u00e9mu. Odvozen\u00fd regresn\u00ed model pro parametr Grd umo\u017e\u0148uje r\u00e1mcov\u011b kvantifikovat vliv tohoto opat\u0159en\u00ed na hydrologick\u00fd syst\u00e9m povod\u00ed. Podobn\u011b p\u0159i sn\u00ed\u017een\u00ed hodnoty CN zm\u011bnou land use a\u00a0zpomalen\u00edm odtoku souborem revitaliza\u010dn\u00edch opat\u0159en\u00ed, kter\u00e9 sni\u017euj\u00ed pr\u016fm\u011brn\u00fd sklon \u0159\u00ed\u010dn\u00ed s\u00edt\u011b, dojde ke zv\u00fd\u0161en\u00ed retence v\u00a0povod\u00ed, kter\u00e1 je charakterizov\u00e1na hodnotou parametru Spa. Odvozen\u00e9 regresn\u00ed vztahy je mo\u017en\u00e9 alternativn\u011b stanovit souhrnnou region\u00e1ln\u00ed kalibrac\u00ed modelu Bilan podle postupu [7].<\/p>\n

Z\u00e1v\u011br<\/h2>\n

V\u00a0p\u0159edlo\u017een\u00e9 studii je navr\u017een metodick\u00fd postup, kter\u00fd d\u00e1le umo\u017e\u0148uje r\u00e1mcov\u011b kvantifikovat vliv p\u0159\u00edrod\u011b bl\u00edzk\u00fdch reten\u010dn\u00edch opat\u0159en\u00ed na hydrologick\u00fd syst\u00e9m povod\u00ed. Navr\u017een\u00fd postup umo\u017e\u0148uje stanovit hodnoty parametr\u016f modelu Bilan pro povod\u00ed, jeho\u017e hydrologick\u00fd syst\u00e9m je ovlivn\u011bn plo\u0161n\u00fdmi a\u00a0liniov\u00fdmi p\u0159\u00edrod\u011b bl\u00edzk\u00fdmi reten\u010dn\u00edmi opat\u0159en\u00edmi. Na v\u00fdsledky studie bude navazovat posouzen\u00ed vlivu souboru reten\u010dn\u00edch opat\u0159en\u00ed kontinu\u00e1ln\u00ed simulac\u00ed hydrologick\u00e9 bilance modelem Bilan v\u00a0povod\u00edch LAPV.<\/p>\n

Pod\u011bkov\u00e1n\u00ed<\/h3>\n

Tento \u010dl\u00e1nek vznikl v\u00a0r\u00e1mci \u0159e\u0161en\u00ed projektu Mo\u017enosti kompenzace negativn\u00edch dopad\u016f klimatick\u00e9 zm\u011bny na z\u00e1sobov\u00e1n\u00ed vodou a\u00a0ekosyst\u00e9my vyu\u017eit\u00edm lokalit vhodn\u00fdch pro akumulaci povrchov\u00fdch vod (TA04020501), kter\u00fd je spolufinancov\u00e1n Technologickou agenturou \u010cesk\u00e9 republiky.<\/p>\n","protected":false},"excerpt":{"rendered":"

In the presented study we tested the selected sets of linear and nonlinear regression models, that describe the relationships between the selected parameters of hydrological model Bilan, which were estimated using the meteorological, hydrological series, and between the retention characteristics estimated using the selected geomorphological patterns of the river basin.<\/p>\n","protected":false},"author":8,"featured_media":2670,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[86],"tags":[679,682,680,678,681],"coauthors":[598,607,29,34],"class_list":["post-2723","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-hydraulics-hydrology-and-hydrogeology","tag-gaussian-process","tag-geomorphological-characteristics","tag-linear-model","tag-regionalization","tag-regression"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/2723","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/comments?post=2723"}],"version-history":[{"count":2,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/2723\/revisions"}],"predecessor-version":[{"id":30365,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/2723\/revisions\/30365"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media\/2670"}],"wp:attachment":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media?parent=2723"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/categories?post=2723"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/tags?post=2723"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/coauthors?post=2723"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}