Vízügyi Közlemények, 1999 (81. évfolyam)
2. füzet - Stanescu, V-A.-Ungureanu, V.-Domokos M.: A Duna-vízgyűjtő regionalizálása az évi nagyvízhozamok eloszlásfüggvényeinek becsléséhez
A Duna-vízgyűjtő regionalizálása az évi nagyvízhozamok... 257 bution functions of the peak discharge modules AT,-, as defined by Eq. (3), were produced for each of the 176 gauging sections of the Danube Catchment with an extension of 817,000 km 2 (in Eq. (3) Q g a j is the symbol of the peak discharge of the ith year, Q g m is the mean value of such peak discharges and C v the variation coefficient thereof). After reading from these distribution functions the values of the module quantiles K p (p being the symbol of the probability of exceedance), a group of relationships K p i=fl,C v) could be plotted for each section (p,= 1%, 2%, ..., 95%). On the basis of the clustering of the latter relationships - as well as of various physiogeographical considerations - five regions of the Danube Catchment could be identified (Fig. I), each of them acceptably homogeneous from the viewpoint of the probability distribution functions of annual peak discharges. For each of the five regions, the series of the relationships K=(C V ) is graphically displayed in Fig. 2 , while their numerical formulae are listed in subchapter 4.1 of the paper. Fig. 3 shows that — as an example — the functions K w=flC v), like the functions of the other quantiles, significantly differ from each other, supporting the goodness of the definition of regions in Fig. I. With the help of these empirical reationships K p=f(C v), the annucal peak discharge Q g p of exceedance p (i.e., the p-quantile of annual peak discharges) can be estimated, by adopting Eq. (4), for any ungauged river section of the Danube Catchment, in which the average value of annual peak discharges Q g m and their variation coefficient C v is known (e.g., from other nomograms). For each of the 176 distribution functions of annual peak discharges, three types of fitting distribution functions (Pearson III, Krickij-Menkelj-type and lognormal) as well as the fitting errors e(p,C v) of the latter from the empirical ones (as defined in Eq. (7)) were determined and listed in Table IV. From the latter one can conclude that (a) the largest errors arise in the region No. 1. (Alps); (b) the error e monotonously increases with C v; (c) the smallest errors generally belong to the fitting distribution function of type Pearson III. In the second part of the investigation ("micro-regionalization") regression relationships according to Eq. (6) were produced between the specific yield for the five hydrologie regions of the national area of Romania (237,500 km 2, 98% of which is lying in the Danube Catchment) q 4j t=q q JA [1/s] and various physiogeographic characteristics X t of the catchments. As it is shown in Table V, the correlation q g [=j(A,H,B/L) proved to be the closest one for most catchments of the five regions, where A is the catchment area [km-], H its altitude above sea level [m], В its average width [km] and L the length of its main watercourse [km]. The constants and estimation errors of the regression relationships (6) of this optimal combination of parameters, depending on the probability of exceedance p, are listed in Table VI for region No. I ., i.e. for the Romanian part of the Tisza Catchment, as a picked-out example. The regression relationships determined as a result of "micro-regionalization" can be used for estimating the quantiles of specific yield, q g p in ungauged river sections of the Danube Catchment, for which the values of the catchment parameters A, H, В and L are known. * * * Regionalisierung des Donaueinzugsgebietes für die Schätzung der Verteilungsfunktionen der jährlichen Hoch Wasserabflüsse von Prof.Dr.-Ing. Viorel-Alexandru STÁNESCU, Dipl.-Ing. Valentina UNGUREANU und Dr. -Ing. Dipl.-Math. Miklós DOMOKOS Eines der Projekte der im Rahmen des IHP/UNESCO, geführten hydrologischen Zusammenarbeit der 13 Donauländer hatte zum Zweck, aus den von 176 Pegelquerschnitten des Donaueinzugsgebietes (Bild l) zur Verfügung stehenden, genügend langen und zuverlässigen Reihen der jährlichen Hochwasserabflüsse (Tabellen I, II, III), unter rumänischer Koordinierung, regionale empirische Beziehungen zu ermitteln, mit deren Hilfe die Quantile der jährlichen Hochwässer bzw. Hochwasserspenden fur die Gewässerquerschnitte mit Datenmangel des Donaueinzugsgebietes abgeschätzt werden können.
