![\"\"](\"http:\/\/www.vtei.cz\/wp-content\/uploads\/2018\/02\/ilustracni-fotografie-3_edit.jpg\")
<\/a>\n\u010cl\u00e1nek p\u0159edstavuje klasifikaci mal\u00fdch povod\u00ed v\u00a0\u010cesk\u00e9 republice pro pot\u0159eby jejich rozd\u011blen\u00ed z\u00a0hlediska potenci\u00e1ln\u00ed hydrologick\u00e9 odezvy na n\u00e1vrhov\u00e9 sr\u00e1\u017eky. Pro celkem 7 739 povod\u00ed IV. \u0159\u00e1du s\u00a0plochou nad 0,5\u2008km2<\/sup> byly ur\u010deny hlavn\u00ed charakteristiky ovliv\u0148uj\u00edc\u00ed hydrologickou reakci dan\u00e9ho povod\u00ed, jako je velikost povod\u00ed, uspo\u0159\u00e1d\u00e1n\u00ed a\u00a0hustota \u0159\u00ed\u010dn\u00ed s\u00edt\u011b, koeficienty prot\u00e1hlosti, d\u00e1le pak pr\u016fm\u011brn\u00e1 nadmo\u0159sk\u00e1 v\u00fd\u0161ka a\u00a0pr\u016fm\u011brn\u00fd sklon. P\u016fdn\u00ed pokryv, reten\u010dn\u00ed kapacita a\u00a0hydropedologick\u00e9 vlastnosti jsou zahrnuty v\u00a0pr\u016fm\u011brn\u00e9m \u010d\u00edsle odtokov\u00e9 k\u0159ivky (CN). Z\u00a0morfologick\u00fdch parametr\u016f, kter\u00e9 maj\u00ed vliv na rychl\u00fd odtok, byla vypo\u010dtena hodnota pr\u016fm\u011brn\u00e9 d\u00e9lky odtokov\u00e9 dr\u00e1hy do nejbli\u017e\u0161\u00edho vodn\u00edho toku a\u00a0nejdel\u0161\u00ed odtokov\u00e1 dr\u00e1ha po uz\u00e1v\u011brov\u00fd profil. Z\u00e1visl\u00e9 parametry byly vy\u0159azeny a\u00a0zbylo tak p\u011bt z\u00e1kladn\u00edch parametr\u016f pro n\u00e1slednou klasifikaci. Rozhoduj\u00edc\u00ed pro klasifikaci byla pouze bl\u00edzkost parametr\u016f dan\u00fdch povod\u00ed nikoli jejich poloha. Postupn\u011b bylo testov\u00e1no seskupov\u00e1n\u00ed do 3 a\u017e 7 t\u0159\u00edd. V\u00a0p\u0159\u00edpad\u011b vy\u0161\u0161\u00edho po\u010dtu t\u0159\u00edd ji\u017e nedoch\u00e1zelo k\u00a0vytvo\u0159en\u00ed jasn\u00fdch skupin. Z\u00a0fin\u00e1ln\u00edch sedmi t\u0159\u00edd dv\u011b nebyly d\u00e1le hodnoceny kv\u016fli jejich relativn\u011b mal\u00e9mu zastoupen\u00ed (m\u00e9n\u011b ne\u017e 10\u00a0% plochy) a\u00a0extremit\u011b specifick\u00fdch parametr\u016f, nej\u010dast\u011bji se jedn\u00e1 o\u00a0tvarov\u011b ovlivn\u011bn\u00e1 a\u00a0\u010dasto velmi mal\u00e1 zbytkov\u00e1 mezipovod\u00ed. V\u00fdsledkem tak je 5 t\u0159\u00edd povod\u00ed v\u00a0r\u00e1mci \u010cR, kter\u00e1 pokr\u00fdvaj\u00ed p\u0159es 90\u00a0% \u00fazem\u00ed. Reprezentativn\u00ed povod\u00ed ka\u017ed\u00e9 t\u0159\u00eddy bylo v\u00a0navazuj\u00edc\u00edch prac\u00edch detailn\u011b modelov\u00e1no v\u00a0modelech MikeSHE, SMODERP a\u00a0HEC-HMS.<\/p>\n Hospoda\u0159en\u00ed s\u00a0vodou v\u00a0krajin\u011b je podm\u00edn\u011bno objemem vyu\u017eiteln\u00fdch vodn\u00edch zdroj\u016f, nem\u00e9n\u011b v\u00fdznamn\u011b v\u0161ak i\u00a0\u010dasovou a\u00a0prostorovou distribuc\u00ed jednotliv\u00fdch slo\u017eek hydrologick\u00e9 bilance, p\u0159edev\u0161\u00edm sr\u00e1\u017eek jako\u017eto prakticky jedin\u00e9 p\u0159\u00edjmov\u00e9 slo\u017eky pro mal\u00e1 povod\u00ed. Podrobnou \u010dasovou distribuci \u0161estihodinov\u00fdch n\u00e1vrhov\u00fdch hyetogram\u016f ned\u00e1vno zpracoval M\u00fcller [1], v\u00a0hrub\u0161\u00edm rozli\u0161en\u00ed hodinov\u00e9ho kroku uvedla t\u0159i typick\u00e9 hyetogramy Kulasov\u00e1 [2] a\u00a0ji\u017e klasickou pr\u00e1ci pak p\u0159edstavuje anal\u00fdza n\u00e1hradn\u00edch maxim\u00e1ln\u00edch intenzit kr\u00e1tkodob\u00fdch de\u0161\u0165\u016f publikovan\u00e1 Truplem [3]. Prostorov\u00e1 variabilita je v\u00a0\u010desk\u00fdch podm\u00ednk\u00e1ch popisov\u00e1na m\u00e9n\u011b, p\u0159\u00edklady uv\u00e1d\u00ed Kavka [4] na z\u00e1klad\u011b dat \u0160amaje [5] nebo Kulasov\u00e1\u00a0[4].<\/p>\n Na formov\u00e1n\u00ed odtoku z\u00a0konkr\u00e9tn\u00edho povod\u00ed maj\u00ed krom\u011b sr\u00e1\u017eky vliv charakteristiky vlastn\u00edho povod\u00ed. Mezi ty nejz\u00e1sadn\u011bj\u0161\u00ed pat\u0159\u00ed vyu\u017eit\u00ed \u00fazem\u00ed (p\u016fdn\u00ed pokryv) a\u00a0infiltra\u010dn\u00ed vlastnosti p\u016fd. Kombinace t\u011bchto dvou charakteristik z\u00e1sadn\u011b ur\u010duje m\u00edru potenci\u00e1ln\u00edho zachycen\u00ed sr\u00e1\u017ekov\u00e9 vody.<\/p>\n Dal\u0161\u00edm faktorem, kter\u00fd ovliv\u0148uje hydrologick\u00e9 procesy, jsou sklonov\u00e9 a\u00a0morfologick\u00e9 charakteristiky. Morfologii povod\u00ed lze charakterizovat celou \u0159adou tvarov\u00fdch koeficient\u016f, jedn\u00edm z\u00a0nich je nap\u0159\u00edklad koeficient prot\u00e1hlosti. V\u00a0mal\u00fdch povod\u00edch je p\u0159eva\u017euj\u00edc\u00edm rizikem p\u0159edev\u0161\u00edm rychl\u00fd odtok z\u00a0intenzivn\u00edch p\u0159\u00ed\u010dinn\u00fdch sr\u00e1\u017eek. Z\u00a0hlediska formov\u00e1n\u00ed povrchov\u00e9ho odtoku spojen\u00e9ho s\u00a0mo\u017en\u00fdm erozn\u00edm ohro\u017een\u00edm je pak nutn\u00e9 do charakteristiky konkr\u00e9tn\u00edho povod\u00ed zahrnout i\u00a0dr\u00e1hy odtoku mimo vodn\u00ed toky (d\u00e9lku drah plo\u0161n\u00e9ho odtoku). Z\u00a0hlediska formov\u00e1n\u00ed korytov\u00e9ho odtoku je pak d\u016fle\u017eit\u00e1 nap\u0159. hustota \u0159\u00ed\u010dn\u00ed s\u00edt\u011b. Nem\u00e9n\u011b d\u016fle\u017eit\u00fdm parametrem je pak pom\u011br d\u00e9lky vodn\u00edch tok\u016f v\u016f\u010di plo\u0161e povod\u00ed, tzv. hustota \u0159\u00ed\u010dn\u00ed s\u00edt\u011b.<\/p>\n Motivac\u00ed pro ur\u010den\u00ed t\u011bchto charakteristik je n\u00e1sledn\u00e1 kategorizace povod\u00ed IV.\u00a0\u0159\u00e1du. U\u00a0vz\u00e1jemn\u011b podobn\u00fdch povod\u00ed lze o\u010dek\u00e1vat i\u00a0podobnou reakci na stejnou p\u0159\u00ed\u010dinnou sr\u00e1\u017eku. Z\u00a0takto vytvo\u0159en\u00fdch kategori\u00ed je pak mo\u017en\u00e9 vybrat reprezentativn\u00ed z\u00e1stupce pro hodnocen\u00ed vlivu charakteristik povod\u00ed na odtokov\u00e9 charakteristiky, jako je odtokov\u00e1 v\u00fd\u0161ka, kulmina\u010dn\u00ed pr\u016ftok nebo doba kulminace. To je zejm\u00e9na v\u00fdhodn\u00e9 p\u0159i pou\u017eit\u00ed v\u00fdpo\u010detn\u011b n\u00e1ro\u010dn\u011bj\u0161\u00edch model\u016f, kter\u00e9 znemo\u017e\u0148uj\u00ed modelov\u00e1n\u00ed rozs\u00e1hl\u00e9ho souboru sr\u00e1\u017eko-odtokov\u00fdch sc\u00e9n\u00e1\u0159\u016f. Naopak p\u0159i pou\u017eit\u00ed jednodu\u0161\u0161\u00edch konceptu\u00e1ln\u00edch hydrologick\u00fdch model\u016f je mo\u017en\u00e9 u\u00a0dan\u00fdch kategori\u00ed povod\u00ed vyhodnotit statistick\u00e9 rozd\u011blen\u00ed popisn\u00fdch charakteristik a\u00a0to pou\u017e\u00edt p\u0159i stochastick\u00e9m modelov\u00e1n\u00ed hydrologick\u00e9 odezvy pro velk\u00e9 mno\u017estv\u00ed kombinac\u00ed charakteristik povod\u00ed a\u00a0z\u00e1t\u011b\u017eov\u00e9 sr\u00e1\u017eky. Pou\u017eit\u00ed obou t\u011bchto p\u0159\u00edstup\u016f a\u00a0v\u00fdb\u011br jejich v\u00fdsledk\u016f je prezentov\u00e1n na jin\u00e9m m\u00edst\u011b v\u00a0tomto \u010dasopise. N\u00e1sleduj\u00edc\u00ed \u010d\u00e1sti tohoto p\u0159\u00edsp\u011bvku se soust\u0159ed\u00ed pouze na popis proveden\u00e9 klasifikace povod\u00ed a\u00a0v\u00fdb\u011br reprezentativn\u00edch z\u00e1stupc\u016f.<\/p>\n Celkem bylo vy\u010d\u00edsleno 15 z\u00e1kladn\u00edch charakteristik povod\u00ed IV. \u0159\u00e1du. Celkov\u011b mohlo b\u00fdt hodnoceno 8 842 povod\u00ed IV. \u0159\u00e1du. Z\u00a0hodnocen\u00ed byla vy\u0159azena velmi mal\u00e1 povod\u00ed s\u00a0velikost\u00ed pod 0,5\u2008km2<\/sup>, kter\u00e1 v\u011bt\u0161inou tvo\u0159\u00ed zbytkov\u00e1 mezipovod\u00ed, a\u00a0kv\u016fli jejich mal\u00e9 velikosti nab\u00fdvaj\u00ed \u010dasto jejich charakteristiky extr\u00e9mn\u00edch hodnot, nebo ztr\u00e1c\u00ed re\u00e1ln\u00fd v\u00fdznam. Z\u00a0hlediska hodnocen\u00ed cel\u00e9 \u010cR se jedn\u00e1 o\u00a0nev\u00fdznamn\u00e9 procento plochy. Pro zbyl\u00fdch 7 739 povod\u00ed s\u00a0plochou nad 0,5\u2008km2<\/sup> byly ur\u010deny charakteristiky uveden\u00e9 v\u00a0tabulce\u00a01<\/em>.<\/p>\n N\u00e1sleduj\u00edc\u00edm krokem bylo kritick\u00e9 zhodnocen\u00ed a\u00a0redukce po\u010dtu parametr\u016f kv\u016fli zpr\u016fhledn\u011bn\u00ed a\u00a0jednozna\u010dnosti klasifika\u010dn\u00ed procedury. Hlavn\u00edm c\u00edlem bylo zachovat pouze parametry, kter\u00e9 jsou v\u00fdznamn\u00e9 z\u00a0hlediska potenci\u00e1ln\u00ed hydrologick\u00e9 odezvy povod\u00ed, a\u00a0z\u00e1rove\u0148 parametry, mezi kter\u00fdmi nen\u00ed v\u00fdznamn\u00e1 vz\u00e1jemn\u00e1 korelace, a\u00a0lze je tedy pova\u017eovat za vz\u00e1jemn\u011b nez\u00e1visl\u00e9. Nap\u0159\u00edklad z\u00a0charakteristik nadmo\u0159sk\u00e9 v\u00fd\u0161ky je vzhledem k\u00a0velikosti povod\u00ed IV. \u0159\u00e1du zcela posta\u010duj\u00edc\u00ed zachovat pouze st\u0159edn\u00ed hodnotu. V\u00a0p\u0159\u00edpad\u011b p\u0159eva\u017euj\u00edc\u00ed sm\u011brov\u00e9 orientace svah\u016f se nejedn\u00e1 o\u00a0hydrologicky v\u00fdznamn\u00fd parametr pro reakci povod\u00ed na kr\u00e1tkodob\u00e9 de\u0161t\u011b, pou\u017e\u00edv\u00e1na je sp\u00ed\u0161e u\u00a0studi\u00ed zohled\u0148uj\u00edc\u00edch t\u00e1n\u00ed sn\u011bhov\u00e9 pokr\u00fdvky. Koeficient reli\u00e9fu je z\u00e1visl\u00fd na pr\u016fm\u011brn\u00e9 nadmo\u0159sk\u00e9 v\u00fd\u0161ce a\u00a0hustota \u0159\u00ed\u010dn\u00ed s\u00edt\u011b m\u016f\u017ee b\u00fdt z\u00e1stupnou veli\u010dinou pro celkovou d\u00e9lku vodn\u00edch tok\u016f. Po kritick\u00e9m zhodnocen\u00ed bylo takto ponech\u00e1no pro dal\u0161\u00ed anal\u00fdzy osm charakteristik, vy\u0159azen\u00e9 jsou v\u00a0tabulce 1<\/em> uvedeny kurz\u00edvou.<\/p>\n Vybran\u00e9 charakteristiky byly vy\u010d\u00edsleny s\u00a0vyu\u017eit\u00edm n\u00e1sleduj\u00edc\u00edch zdroj\u016f a\u00a0postup\u016f:<\/p>\n Tabulka 2<\/em> uv\u00e1d\u00ed korela\u010dn\u00ed koeficienty mezi zbyl\u00fdmi osmi charakteristikami povod\u00ed. V\u00fdznamnou korelaci lze nal\u00e9zt mezi parametrem sklonu a\u00a0hodnotou CN (-0,76), nadmo\u0159skou v\u00fd\u0161kou a\u00a0CN (-0,57) a\u00a0pr\u016fm\u011brn\u00fdm sklonem a\u00a0nadmo\u0159skou v\u00fd\u0161kou (0,49). To odpov\u00edd\u00e1 skute\u010dnosti, \u017ee ve vy\u0161\u0161\u00edch nadmo\u0159sk\u00fdch v\u00fd\u0161k\u00e1ch, stejn\u011b jako na prud\u0161\u00edch svaz\u00edch, se \u010dast\u011bji vyskytuj\u00ed lesn\u00ed porosty s\u00a0n\u00edzkou hodnotou CN. Naopak t\u00e9m\u011b\u0159 nez\u00e1visl\u00e9 jsou v\u0161echny parametry s\u00a0plochou povod\u00ed, krom\u011b Lmax.s, kter\u00e1 je plochou definov\u00e1na. Podobn\u011b \u017e\u00e1dn\u00e1 korelace nebyla nalezena mezi sklonem povod\u00ed, hustotou \u0159\u00ed\u010dn\u00ed s\u00edt\u011b a\u00a0koeficientem tvaru povod\u00ed \u03b1 (korela\u010dn\u00ed koeficient kolem +\/- 0,05).<\/p>\n Pro klasifikaci 7 739 povod\u00ed IV. \u0159\u00e1du byl ve fin\u00e1le pou\u017eit v\u00fdb\u011br p\u011bti charakteristik z\u00a0v\u00fd\u0161e uveden\u00e9ho seznamu osmi posuzovan\u00fdch. T\u0159i byly dodate\u010dn\u011b vy\u0159azeny z\u00a0r\u016fzn\u00fdch d\u016fvod\u016f. I\u00a0kdy\u017e se parametr plochy povod\u00ed ukazuje jako nez\u00e1visl\u00fd, neexistuje ve skute\u010dnosti \u017e\u00e1dn\u00fd p\u0159\u00edm\u00fd mechanismus, kter\u00fdm by mohl ovlivnit modelovanou odezvu povod\u00ed. Podobn\u011b parametr pr\u016fm\u011brn\u00e9 nadmo\u0159sk\u00e9 v\u00fd\u0161ky byl vynech\u00e1n, nebo\u0165 se jedn\u00e1 o\u00a0charakteristiku ovliv\u0148uj\u00edc\u00ed sp\u00ed\u0161e parametry z\u00e1t\u011b\u017eov\u00e9 sr\u00e1\u017eky. Variabilita sr\u00e1\u017eek v\u0161ak byla v\u00a0navazuj\u00edc\u00edm modelov\u00e1n\u00ed zohledn\u011bna v\u00fdb\u011brem sr\u00e1\u017ekov\u00fdch sc\u00e9n\u00e1\u0159\u016f. Nakonec byl z\u00a0uva\u017eovan\u00fdch charakteristik vy\u0159azen pr\u016fm\u011brn\u00fd sklon povod\u00ed pro zna\u010dnou korelaci s\u00a0hodnotami CN. Pro klasifikaci charakteristick\u00fdch skupin povod\u00ed tedy bylo zvoleno t\u011bchto p\u011bt charakteristik:<\/p>\n Samotn\u00e1 klasifikace povod\u00ed na z\u00e1klad\u011b vybran\u00fdch charakteristik byla provedena v\u00a0prost\u0159ed\u00ed ArcGIS n\u00e1strojem Grouping Analysis s\u00a0vyu\u017eit\u00edm metody K-means. Nebyla zad\u00e1na \u017e\u00e1dn\u00e1 omezen\u00ed uva\u017euj\u00edc\u00ed polohu povod\u00ed (No Spatial Constraints), tedy povod\u00ed ze stejn\u00e9 t\u0159\u00eddy spolu nemus\u00ed nutn\u011b sousedit a\u00a0rozhoduj\u00edc\u00ed pro p\u0159i\u0159azen\u00ed do t\u0159\u00eddy je pouze bl\u00edzkost jejich parametr\u016f. Postupn\u011b bylo testov\u00e1no seskupov\u00e1n\u00ed do 3 a\u017e 7 t\u0159\u00edd. Pro ka\u017edou t\u0159\u00eddu byly spo\u010dteny popisn\u00e9 statistiky pou\u017eit\u00fdch charakteristik povod\u00ed. Pr\u016fm\u011brn\u00e9 hodnoty pro ka\u017edou t\u0159\u00eddu pak byly standardizov\u00e1ny vzhledem k\u00a0pr\u016fm\u011bru a\u00a0odchylce cel\u00e9ho souboru a\u00a0vyneseny v\u00a0podob\u011b paraleln\u00edho grafu. V\u00fdvoj paraleln\u00edho grafu p\u0159i postupn\u00e9m zvy\u0161ov\u00e1n\u00ed po\u010dtu t\u0159\u00edd ukazuje obr. 1<\/em>. Barvy ve v\u0161ech grafech a\u00a0zna\u010den\u00ed v\u00a0n\u00e1sleduj\u00edc\u00edm textu odpov\u00eddaj\u00ed paraleln\u00edmu grafu pro maxim\u00e1ln\u00ed po\u010det sedmi t\u0159\u00edd.<\/p>\n P\u0159i klasifikaci povod\u00ed do t\u0159\u00ed t\u0159\u00edd se vyd\u011bluje v\u00fdrazn\u00e1 t\u0159\u00edda TR4, kter\u00e9 dominuj\u00ed povod\u00ed s\u00a0velmi vysokou specifickou d\u00e9lkou nejdel\u0161\u00ed odtokov\u00e9 dr\u00e1hy Lmax.s. Tato t\u0159\u00edda je v\u0161ak pom\u011brn\u011b m\u00e1lo po\u010detn\u00e1 a\u00a0zbyl\u00e9 dv\u011b TR2 a\u00a0TR7 se li\u0161\u00ed zejm\u00e9na hodnotou CN. P\u0159i zv\u00fd\u0161en\u00ed po\u010dtu t\u0159\u00edd na 4 se rozpad\u00e1 t\u0159\u00edda TR7 a\u00a0vznik\u00e1 nov\u00e1 t\u0159\u00edda TR1, rozhoduj\u00edc\u00edm krit\u00e9riem se zde zd\u00e1 b\u00fdt tvar povod\u00ed vyj\u00e1d\u0159en\u00fd koeficientem \u03b1. P\u0159i dal\u0161\u00edm zv\u00fd\u0161en\u00ed po\u010dtu t\u0159\u00edd na 5 se d\u00e1le rozpad\u00e1 TR7 a\u00a0vznik\u00e1 nov\u00e1 t\u0159\u00edda TR5, v\u00fdrazn\u00fdm rozli\u0161ovac\u00edm znakem je d\u00e9lka st\u0159edn\u00edch drah plo\u0161n\u00e9ho odtoku OFLen.mean. Zv\u00fd\u0161en\u00edm po\u010dtu t\u0159\u00edd na 6 se u\u017e tak relativn\u011b m\u00e1lo po\u010detn\u00e1 skupina TR4 rozpad\u00e1 a\u00a0vyd\u011bluje se z\u00a0n\u00ed velmi mal\u00e1 t\u0159\u00edda TR3 s\u00a0extr\u00e9mn\u00edmi hodnotami hustoty \u0159\u00ed\u010dn\u00ed s\u00edt\u011b HRS a\u00a0specifickou d\u00e9lkou nejdel\u0161\u00ed odtokov\u00e9 dr\u00e1hy Lmax.s. \u010citelnost v\u00fdstup\u016f klasifika\u010dn\u00edho n\u00e1stroje se zastavila p\u0159i nav\u00fd\u0161en\u00ed po\u010dtu t\u0159\u00edd na 7, kdy posledn\u00ed nov\u00e1 t\u0159\u00edda TR6 vznik\u00e1 pravd\u011bpodobn\u011b vyd\u011blen\u00edm z\u00a0TR2. Tyto dv\u011b t\u0159\u00eddy maj\u00ed t\u00e9m\u011b\u0159 shodn\u00e9 parametry a\u017e na inverzn\u00ed vztah hustoty \u0159\u00ed\u010dn\u00ed s\u00edt\u011b HRS a\u00a0pr\u016fm\u011brn\u00e9 st\u0159edn\u00ed dr\u00e1hy plo\u0161n\u00e9ho odtoku OFLen.mean.<\/p>\n Pro sedm vznikl\u00fdch t\u0159\u00edd byly vy\u010d\u00edsleny \u010detnosti jejich zastoupen\u00ed. Kv\u016fli mal\u00e9 \u010detnosti (57, resp. 613 povod\u00ed ze\u00a07 739) nebo velk\u00e9mu pod\u00edlu nestandardn\u00edch povod\u00ed (zpravidla mezipovod\u00ed nebo povod\u00ed odvod\u0148ovac\u00edch kan\u00e1l\u016f) bylo rozhodnuto vypustit z\u00a0klasifikace t\u0159\u00eddy TR3 a\u00a0TR4 a povod\u00ed spadaj\u00edc\u00ed do t\u011bchto t\u0159\u00edd za\u0159adit do nehodnocen\u00fdch povod\u00ed k\u00a0povod\u00edm do 0,5\u2008km2<\/sup>. V\u00fdsledn\u00fd po\u010det t\u0159\u00edd se tedy ust\u00e1lil na 5, kter\u00e9 zastupuj\u00ed p\u0159es 90\u00a0% plochy \u010cR, a\u00a0skupina nehodnocen\u00fdch povod\u00ed (mal\u00e1 povod\u00ed a\u00a0skupiny TR3 a\u00a0TR4).<\/p>\n V\u00fdsledn\u00e9 t\u0159\u00eddy byly pro dal\u0161\u00ed \u0159e\u0161en\u00ed p\u0159e\u010d\u00edslov\u00e1ny na souvislou \u0159adu 1\u20135 se zachov\u00e1n\u00edm vzestupn\u00e9ho po\u0159ad\u00ed z\u00a0p\u016fvodn\u00ed klasifikace (tedy TR1-2-5-6-7 \u2192 TR1-2-3-4-5) a\u00a0-1 pro nehodnocen\u00e1 povod\u00ed. D\u016fvodem p\u0159e\u010d\u00edslov\u00e1n\u00ed bylo automatick\u00e9 \u010d\u00edslov\u00e1n\u00ed skupin n\u00e1strojem Grouping Analysis, bez mo\u017enosti ovlivnit po\u0159ad\u00ed. Zastoupen\u00ed v\u00fdsledn\u00fdch t\u0159\u00edd povod\u00ed IV. \u0159\u00e1du na \u00fazem\u00ed \u010cR je patrn\u00e9 na obr. 2<\/em>.<\/p>\n Jednotliv\u00e9 t\u0159\u00eddy reprezentuj\u00ed skupinu povod\u00ed s\u00a0bl\u00edzk\u00fdmi charakteristikami, podle nich\u017e lze t\u0159\u00eddy popsat n\u00e1sledovn\u011b:<\/p>\n Pro ka\u017edou t\u0159\u00eddu povod\u00ed byla ur\u010dena poloha t\u011b\u017ei\u0161t\u011b v\u00a0parametrick\u00e9m prostoru (dan\u00e1 sou\u0159adnicemi v\u00a0tabulce 3<\/em>) a\u00a0variabilita parametr\u016f uvnit\u0159 t\u0159\u00eddy. U\u00a0ka\u017ed\u00e9ho ze 7 739 povod\u00ed byla pot\u00e9 spo\u010dtena jeho standardizovan\u00e1 odlehlost od t\u011b\u017ei\u0161t\u011b jemu p\u0159\u00edslu\u0161ej\u00edc\u00ed t\u0159\u00eddy. Se\u0159azen\u00edm povod\u00ed podle t\u00e9to odlehlosti bylo mo\u017en\u00e9 vyhledat vhodn\u00e9 kandid\u00e1ty nejl\u00e9pe reprezentuj\u00edc\u00ed jednotliv\u00e9 t\u0159\u00eddy. Po podrobn\u011bj\u0161\u00edm studiu mapov\u00fdch podklad\u016f a\u00a0vylou\u010den\u00ed pro modelov\u00e1n\u00ed nevhodn\u00fdch kandid\u00e1t\u016f byla jako reprezentativn\u00ed povod\u00ed zastupuj\u00edc\u00ed jednotliv\u00e9 t\u0159\u00eddy vybr\u00e1na povod\u00ed uveden\u00e1 v\u00a0tabulce 4<\/em>. Jejich polohu v\u00a0r\u00e1mci \u010cR ukazuje obr. 3<\/em>.<\/p>\n Pomoc\u00ed shlukov\u00e9 anal\u00fdzy charakteristik povod\u00ed IV. \u0159\u00e1du bylo ur\u010deno p\u011bt p\u0159eva\u017euj\u00edc\u00edch skupin povod\u00ed. Jejich zat\u0159\u00edd\u011bn\u00ed pak umo\u017e\u0148uje v\u00fdb\u011br vhodn\u00fdch z\u00e1stupc\u016f t\u0159\u00edd pro navazuj\u00edc\u00ed hodnocen\u00ed variability hydrologick\u00e9 odezvy povod\u00ed na p\u0159\u00ed\u010dinn\u00e9 sr\u00e1\u017eky. V\u00a0p\u016fvodn\u00edch patn\u00e1cti parametrech byly nalezeny parametry s\u00a0v\u00fdraznou vz\u00e1jemnou korelac\u00ed, co\u017e vedlo k\u00a0redukci jejich po\u010dtu. Jedn\u00edm z\u00a0p\u0159ekvapiv\u00fdch z\u00e1v\u011br\u016f byla velmi siln\u00e1 z\u00e1vislost mezi hodnotou CN a\u00a0pr\u016fm\u011brn\u00fdm sklonem povod\u00ed. Z\u00a0prostorov\u00e9ho rozlo\u017een\u00ed v\u00fdsledn\u00fdch t\u0159\u00edd povod\u00ed vypl\u00fdv\u00e1, \u017ee nen\u00ed mo\u017en\u00e9 podle zvolen\u00fdch krit\u00e9ri\u00ed roz\u010dlenit \u00fazem\u00ed \u010cR do homogenn\u00edch celk\u016f, ale \u017ee se r\u016fzn\u00e9 typy povod\u00ed vyskytuj\u00ed prakticky na cel\u00e9m \u00fazem\u00ed. Je mo\u017en\u00e9 sledovat p\u0159eva\u017euj\u00edc\u00ed skupiny povod\u00ed v\u00a0horsk\u00fdch oblastech, co\u017e je d\u00e1no vy\u0161\u0161\u00ed m\u00edrou zalesn\u011bn\u00ed, a\u00a0t\u00edm p\u00e1dem ni\u017e\u0161\u00ed hodnotou CN. Proveden\u00e1 klasifikace povod\u00ed do t\u0159\u00edd podle podobn\u00fdch charakteristik umo\u017en\u00ed prov\u00e9st n\u00e1vaznou citlivostn\u00ed anal\u00fdzu fyzik\u00e1ln\u00edch hydrologick\u00fdch model\u016f pouze na vybran\u00fdch reprezentativn\u00edch povod\u00edch dan\u00e9 t\u0159\u00eddy. M\u00edsto modelov\u00e1n\u00ed velk\u00e9ho mno\u017estv\u00ed povod\u00ed mohou b\u00fdt pou\u017eiti pouze reprezentativn\u00ed z\u00e1stupci jednotliv\u00fdch t\u0159\u00edd povod\u00ed. V\u00a0p\u0159\u00edpad\u011b konceptu\u00e1ln\u00edch model\u016f pak klasifikace umo\u017en\u00ed stochastick\u00e9 modelov\u00e1n\u00ed d\u00edky odli\u0161n\u00fdm statistick\u00fdm rozd\u011blen\u00edm vstupn\u00edch parametr\u016f u\u00a0jednotliv\u00fdch t\u0159\u00edd povod\u00ed.<\/p>\n Anal\u00fdza charakteristik povod\u00ed a\u00a0jejich klasifikace byly motivov\u00e1ny v\u00fdb\u011brem reprezentativn\u00edch povod\u00ed pro pot\u0159eby projektu NAZV QJ1520265 Vliv variability kr\u00e1tkodob\u00fdch sr\u00e1\u017eek a\u00a0n\u00e1sledn\u00e9ho odtoku v\u00a0mal\u00fdch povod\u00edch \u010cesk\u00e9 republiky na hospoda\u0159en\u00ed s\u00a0vodou v\u00a0krajin\u011b, s\u00a0jeho\u017e podporou vznikla.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":" This article introduces a classification of small watersheds in the Czech Republic in terms of potential hydrological response to a set of design precipitation time series. Watershed classes as well as precipitation time series were derived as a partial result of a three-year project aimed at the design of typical soil conservation measures and small water management structures<\/p>\n","protected":false},"author":8,"featured_media":4401,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[86],"tags":[749,1101,1098,1100],"coauthors":[1066,1065,1068,1067],"class_list":["post-4457","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-hydraulics-hydrology-and-hydrogeology","tag-catchment-characteristics","tag-design-rainfalls","tag-hydrological-modelling","tag-small-watersheds"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/4457","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=4457"}],"version-history":[{"count":2,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/4457\/revisions"}],"predecessor-version":[{"id":30444,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/4457\/revisions\/30444"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media\/4401"}],"wp:attachment":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media?parent=4457"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/categories?post=4457"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/tags?post=4457"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/coauthors?post=4457"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}\u00davod<\/h2>\n
<\/a>\nObr. 1. V\u00fdvoj t\u0159\u00edd povod\u00ed p\u0159i klasifikaci do 3\u20137 shluk\u016f metodou K-means; hodnoty p\u0159edstavuj\u00ed pr\u016fm\u011bry charakteristik pro danou t\u0159\u00eddu standardizovan\u00e9 vzhledem k pr\u016fm\u011bru a sm\u011brodatn\u00e9 odchylce charakteristiky cel\u00e9ho souboru
\nFig. 1. Evolution of watershed classes from three to seven clusters using K-means method; values on the y-axis present the class means of particular characteristic standardized with respect to mean and standard deviation of the whole set<\/h6>\n<\/a>\nObr. 2. Prostorov\u00e9 rozlo\u017een\u00ed p\u011bti t\u0159\u00edd povod\u00ed klasifikovan\u00fdch podle podobnosti charakteristik
\nFig. 2. Regionalization of five watershed classes based on their characteristics similarity<\/h6>\nMetody a\u00a0postupy<\/h2>\n
Ur\u010den\u00ed a\u00a0v\u00fdb\u011br charakteristik povod\u00ed<\/h3>\n
Tabulka 1. Seznam charakteristik povod\u00ed uva\u017eovan\u00fdch pro klasifikaci (kurz\u00edvou uvedeny pozd\u011bji vy\u0159azen\u00e9 veli\u010diny)
\nTable 1. List of catchment characteristics considered for classification (in italic later discarded parameters)<\/h5>\n<\/a>\n\n
Tabulka 2. Korela\u010dn\u00ed matice pro osm vybran\u00fdch charakteristik povod\u00ed
\nTable 2. Correlation matrix for 8 selected catchment characteristics<\/h5>\n<\/a>\nKlasifikace povod\u00ed<\/h3>\n
\n
<\/a>\nObr. 3. Poloha vybran\u00fdch charakteristick\u00fdch povod\u00ed s\u00a0\u010d\u00edslem t\u0159\u00eddy
\nFig. 3. Location of representative catchments with their class number<\/h6>\nV\u00fdsledn\u00e9 t\u0159\u00eddy povod\u00ed<\/h2>\n
Tabulka 3. Absolutn\u00ed pr\u016fm\u011brn\u00e9 hodnoty charakteristik v\u00fdsledn\u00fdch t\u0159\u00edd povod\u00ed
\nTable 3. Absolute mean values of characteristics of final catchments classes<\/h5>\n<\/a>\n\n
Tabulka 4. Reprezentativn\u00ed povod\u00ed pro v\u00fdsledn\u00e9 t\u0159\u00eddy vze\u0161l\u00e9 z\u00a0klasifikace
\nTable 4. Representative catchments for five final classes<\/h5>\n<\/a>\nZ\u00e1v\u011br<\/h2>\n
Pod\u011bkov\u00e1n\u00ed<\/h3>\n