"Va\u0161\u0161ov\u00E1, Darina" . . "Neruda, Roman" . "Jel\u00EDnkov\u00E1, Andrea" . "14"^^ . . . "\u0160rejber, Jan" . . . . "I, S" . "flood forecasting; neural network; Sm\u011Bd\u00E1 catchment; KINFIL"@en . "Possibilities of usage of different hydrological models in the research of rainfall-runoff extremes in small catchments"@en . "CZ - \u010Cesk\u00E1 republika" . "Neruda, Martin" . "1"^^ . "1802-212X" . "Often task for real and scenario prognosis in engineering hydrology is usage of simulation techniques of mathematical models for processes in small catchments. These catchments have often area till 35 km2, their character is subcritical in mountainous and sub mountainous areas (index of torrent KB} 0,1) and often there is not a water stage gauge. Damages in their catchments are enormous and length of torrents is about 35 % (18 000 km) of the total length of small rivers in the Czech Republic (B\u011Blsk\u00FD, 1999). An experimental mountain catchment Sm\u011Bd\u00E1 (profile %22B\u00EDl\u00FD potok%22) in %22Jizersk\u00E9 hory%22 Mountains was chosen as model area for simulation of extreme rainfall.runoff processes of two different models. For evaluation and simulations of important rainfall runoff episodes we have chosen a physical based hydrological 2D model KINFIL and a mathematical %22learning%22 model of application neural networks MANS. Neural network is a mathematical model of non linear functional dependence between inputs and outputs with free parameters (weights), which are made by gradient learning algorithms with much iteration, where calibrating data are run."@en . "VII" . . . . . "\u010Cast\u00FDm \u00FAkolem re\u00E1ln\u00FDch i sc\u00E9n\u00E1\u0159ov\u00FDch progn\u00F3z in\u017Een\u00FDrsk\u00E9 hydrologie je vyu\u017Eit\u00ED simula\u010Dn\u00ED techniky matematick\u00FDch model\u016F pro procesy na mal\u00FDch povod\u00EDch. Tato povod\u00ED m\u00EDvaj\u00ED \u010Dasto plochu do 35 km2, jejich charakter b\u00FDv\u00E1 v horsk\u00FDch a podhorsk\u00FDch oblastech byst\u0159inn\u00FD (index byst\u0159innosti KB} 0,1) a obvykle nejsou vybaveny limnigrafick\u00FDm m\u011B\u0159en\u00EDm. \u0160kody, kter\u00E9 p\u016Fsob\u00ED v povod\u00ED, b\u00FDvaj\u00ED enormn\u00ED a rozsah t\u011Bchto byst\u0159in je asi 35 % (18 000 km) d\u00E9lky drobn\u00FDch vodn\u00EDch tok\u016F \u010Cesk\u00E9 republiky (B\u011Blsk\u00FD, 1999). Proto bylo vybr\u00E1no experiment\u00E1ln\u00ED povod\u00ED byst\u0159iny Sm\u011Bd\u00E9 (UP B\u00EDl\u00FD potok) v Jizersk\u00FDch hor\u00E1ch jako modelov\u00E9 \u00FAzem\u00ED pro simulace extr\u00E9mn\u00EDch sr\u00E1\u017Eko-odtokov\u00FDch proces\u016F dvou odli\u0161n\u00FDch model\u016F. Pro vz\u00E1jemn\u00E9 vyu\u017Eit\u00ED jsme vybrali k simulaci v\u00FDznamn\u00FDch sr\u00E1\u017Eko-odtokov\u00FDch epizod fyzik\u00E1ln\u011B zalo\u017Een\u00FD hydrologick\u00FD 2D model KINFIL a \u010Dist\u011B matematick\u00FD %22u\u010D\u00EDc\u00ED se%22 model aplikace neuronov\u00FDch s\u00EDt\u00ED MANS. Neuronov\u00E1 s\u00ED\u0165 je matematick\u00FDm modelem neline\u00E1rn\u00ED funk\u010Dn\u00ED z\u00E1vislosti mezi vstupy a v\u00FDstupy s voln\u00FDmi parametry (v\u00E1hami), kter\u00E9 se nastavuj\u00ED gradientn\u00EDm u\u010D\u00EDc\u00EDm algoritmem s mnoha iteracemi, b\u011Bhem kter\u00FDch se proch\u00E1zej\u00ED kalibra\u010Dn\u00ED data."@cs . . "89913" . . "6"^^ . "Mo\u017Enosti vyu\u017Eit\u00ED odli\u0161n\u00FDch hydrologick\u00FDch model\u016F v \u0159e\u0161en\u00ED sr\u00E1\u017Eko-odtokov\u00FDch extr\u00E9m\u016F na mal\u00FDch povod\u00EDch"@cs . "Mo\u017Enosti vyu\u017Eit\u00ED odli\u0161n\u00FDch hydrologick\u00FDch model\u016F v \u0159e\u0161en\u00ED sr\u00E1\u017Eko-odtokov\u00FDch extr\u00E9m\u016F na mal\u00FDch povod\u00EDch"@cs . "RIV/44555601:13520/13:43885486" . . "Mo\u017Enosti vyu\u017Eit\u00ED odli\u0161n\u00FDch hydrologick\u00FDch model\u016F v \u0159e\u0161en\u00ED sr\u00E1\u017Eko-odtokov\u00FDch extr\u00E9m\u016F na mal\u00FDch povod\u00EDch" . "Mo\u017Enosti vyu\u017Eit\u00ED odli\u0161n\u00FDch hydrologick\u00FDch model\u016F v \u0159e\u0161en\u00ED sr\u00E1\u017Eko-odtokov\u00FDch extr\u00E9m\u016F na mal\u00FDch povod\u00EDch" . "Possibilities of usage of different hydrological models in the research of rainfall-runoff extremes in small catchments"@en . "13520" . . "2" . "\u010Cast\u00FDm \u00FAkolem re\u00E1ln\u00FDch i sc\u00E9n\u00E1\u0159ov\u00FDch progn\u00F3z in\u017Een\u00FDrsk\u00E9 hydrologie je vyu\u017Eit\u00ED simula\u010Dn\u00ED techniky matematick\u00FDch model\u016F pro procesy na mal\u00FDch povod\u00EDch. Tato povod\u00ED m\u00EDvaj\u00ED \u010Dasto plochu do 35 km2, jejich charakter b\u00FDv\u00E1 v horsk\u00FDch a podhorsk\u00FDch oblastech byst\u0159inn\u00FD (index byst\u0159innosti KB} 0,1) a obvykle nejsou vybaveny limnigrafick\u00FDm m\u011B\u0159en\u00EDm. \u0160kody, kter\u00E9 p\u016Fsob\u00ED v povod\u00ED, b\u00FDvaj\u00ED enormn\u00ED a rozsah t\u011Bchto byst\u0159in je asi 35 % (18 000 km) d\u00E9lky drobn\u00FDch vodn\u00EDch tok\u016F \u010Cesk\u00E9 republiky (B\u011Blsk\u00FD, 1999). Proto bylo vybr\u00E1no experiment\u00E1ln\u00ED povod\u00ED byst\u0159iny Sm\u011Bd\u00E9 (UP B\u00EDl\u00FD potok) v Jizersk\u00FDch hor\u00E1ch jako modelov\u00E9 \u00FAzem\u00ED pro simulace extr\u00E9mn\u00EDch sr\u00E1\u017Eko-odtokov\u00FDch proces\u016F dvou odli\u0161n\u00FDch model\u016F. Pro vz\u00E1jemn\u00E9 vyu\u017Eit\u00ED jsme vybrali k simulaci v\u00FDznamn\u00FDch sr\u00E1\u017Eko-odtokov\u00FDch epizod fyzik\u00E1ln\u011B zalo\u017Een\u00FD hydrologick\u00FD 2D model KINFIL a \u010Dist\u011B matematick\u00FD %22u\u010D\u00EDc\u00ED se%22 model aplikace neuronov\u00FDch s\u00EDt\u00ED MANS. Neuronov\u00E1 s\u00ED\u0165 je matematick\u00FDm modelem neline\u00E1rn\u00ED funk\u010Dn\u00ED z\u00E1vislosti mezi vstupy a v\u00FDstupy s voln\u00FDmi parametry (v\u00E1hami), kter\u00E9 se nastavuj\u00ED gradientn\u00EDm u\u010D\u00EDc\u00EDm algoritmem s mnoha iteracemi, b\u011Bhem kter\u00FDch se proch\u00E1zej\u00ED kalibra\u010Dn\u00ED data." . "RIV/44555601:13520/13:43885486!RIV14-MSM-13520___" . "Studia Oecologica" . . . "[8D55C6C0385C]" . "Kov\u00E1\u0159, Pavel" . .