\n
\n<\/li>\n
\n
\n<\/li>\n
\n
\n<\/li>\n
\n
\n<\/li>\n
Based on existing knowledge and available source materials, the\u00a0following initial hypotheses are formulated and subsequently tested within\u00a0the\u00a0study:<\/p>\n
\n
- \n
- The\u00a0peak discharge of\u00a0the\u00a01903 flood at the\u00a0Opava profile was significantly higher than\u00a0values derived using simpler historical calculation methods and lay approximately within\u00a0the\u00a0range corresponding to extreme discharges (hundreds of\u00a0m\u00b3\u2009\u00b7\u2009s-1<\/sup>).
\n
\n<\/li>\n- The\u00a0combination of\u00a0GIS-based spatial analyses, rainfall\u2013runoff modelling, and\u00a0hydrodynamic modelling (1D\/2D) enables a\u00a0more accurate and physically consistent reconstruction of\u00a0historical floods than\u00a0approaches based solely on\u00a0empirical relationships or simplified hydraulic calculations.<\/li>\n
- The\u00a0results of\u00a0the\u00a0hydrological and hydraulic modelling are broadly consistent with preserved flood marks and the\u00a0documented extent of\u00a0flooding and may therefore be regarded as a\u00a0realistic approximation of\u00a0the\u00a0course of\u00a0the\u00a0flood.
\n
\n<\/li>\n- The\u00a01903 flood represented an\u00a0extreme hydrological event comparable to \u2013 or locally even more severe than\u00a0\u2013 the\u00a0floods of\u00a01997 and 2024 and exceeded commonly considered design discharges.
\n
\n<\/li>\n- During extreme rainfall episodes and under conditions of\u00a0high antecedent catchment saturation, meteorological forcing and the\u00a0morphometric characteristics of\u00a0the\u00a0catchment play a\u00a0dominant role, whereas the\u00a0influence of\u00a0differences in\u00a0land use gradually decreases.
\n
\n<\/li>\n- Uncertainties associated with the\u00a0lack of\u00a0input data can\u00a0be quantified using stochastic approaches (e.g. Monte Carlo simulations), and these uncertainties significantly influence the\u00a0resulting range of\u00a0estimated peak discharges.<\/li>\n<\/ol>\n
The\u00a0formulated objectives, research questions, and hypotheses provide the\u00a0framework for the\u00a0subsequent application of\u00a0modelling tools and the\u00a0interpretation of\u00a0the\u00a0results of\u00a0the\u00a0reconstruction of\u00a0the\u00a01903 flood.<\/p>\n
METEOROLOGICAL SITUATION<\/h2>\n
The\u00a0high level of\u00a0catchment saturation was an\u00a0important factor in\u00a0the\u00a0development of\u00a0the\u00a0July 1903 flood. June 1903\u00a0had already been very wet in\u00a0the\u00a0Odra basin. For example, more than\u00a060\u00a0mm of\u00a0precipitation fell on average during the\u00a0period from 11 to 16\u00a0June alone. As a\u00a0result, catchment saturation was already considerable by 6\u00a0July. Between 6 and 8\u00a0July, a\u00a0wavy cold front extended across Silesia and Moravia further southwards, separating cold air over western Europe from warm air over eastern Europe. In\u00a0the\u00a0Jesen\u00edky region, a\u00a0further approximately 35\u00a0mm of\u00a0precipitation fell during these days, so that by the\u00a0morning of\u00a09\u00a0July the\u00a0saturation of\u00a0the\u00a0local catchments was almost twice the\u00a0normal level for that time of\u00a0year.<\/p>\n
A\u00a0cyclone formed along this frontal boundary over northern Italy as early as 7\u00a0July and subsequently moved north-eastwards along the\u00a0so-called Vb track in\u00a0the\u00a0following days. On 9\u00a0July 1903, it was situated over south-eastern Europe; along its eastern and northern margins, warm and humid air from the\u00a0eastern Mediterranean\u00a0flowed over Moravia and Silesia (Fig.\u00a01<\/em>). A\u00a0northerly wind prevailed in\u00a0the\u00a0lower atmospheric layers and was intensified by the\u00a0strong horizontal pressure gradient between the\u00a0cyclone and an\u00a0anticyclone centred over the\u00a0British Isles. As a\u00a0result, pronounced orographic enhancement of\u00a0precipitation occurred, with the\u00a0precipitation maximum located along the\u00a0northern edge of\u00a0Hrub\u00fd Jesen\u00edk; most of\u00a0the\u00a0area where totals exceeded 100\u00a0mm is drained into the\u00a0Odra basin\u00a0(particularly the\u00a0Opava, B\u011bl\u00e1, and Vidn\u00e1vka catchments) within\u00a0present-day Czech and Polish territory. A\u00a0second anticyclone extending over eastern Europe also played a\u00a0role by preventing the\u00a0displacement of\u00a0the\u00a0precipitation-producing cyclone.<\/p>\n
![\"\"](\"data:image\/svg+xml;base64,PHN2ZyB3aWR0aD0iMSIgaGVpZ2h0PSIxIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg==\")
<\/a>\nFig. 1. Synoptic situation on 9\u00a0July 1903, expressed by the\u00a0height of\u00a0the\u00a0isobaric level of\u00a0850 hPa in\u00a0gpm (white isohypses) and the\u00a0air temperature [\u00b0C] at this level (colour scale); the\u00a0letters indicate the\u00a0cyclone (N) and the\u00a0anticyclone (V) (source:\u00a0Wetterzentrale.de)<\/h6>\n
During this event, 240.2\u00a0mm of\u00a0precipitation \u2013 i.e. 240.2\u00a0litres per m\u00b2 \u2013 was measured at Nov\u00e1 \u010cerven\u00e1 Voda station on 9\u00a0July, a\u00a0total that was not exceeded during the\u00a020th century until July 1997 at Studni\u010dn\u00ed hora in\u00a0Krkono\u0161e. Until the\u00a0flood of\u00a0September 2024, this remained the\u00a0daily precipitation record for Moravia, Bohemian\u00a0Silesia, and Jesen\u00edky. This exceptionally high daily precipitation total is remarkable given the\u00a0low elevation of\u00a0the\u00a0station (310\u00a0m a.s.l.), which is nevertheless situated in\u00a0the\u00a0northern foreland of\u00a0a\u00a0mountain\u00a0ridge whose highest peak, Studni\u010dn\u00ed vrch, reaches 992\u00a0m a.s.l. In\u00a0addition, precipitation totals of\u00a0200\u00a0mm or more were recorded that day at three other stations: 221.0\u00a0mm at Star\u00fd Rejv\u00edz (757\u00a0m a.s.l.), 217.7\u00a0mm at the\u00a0\u0160umn\u00fd potok station (559\u00a0m a.s.l.), and 200\u00a0mm at L\u00e1zn\u011b Jesen\u00edk (625\u00a0m a.s.l.).<\/span><\/p>\n
The\u00a0extraordinary flooding was exceptionally destructive, particularly in\u00a0Jesen\u00edky, where the\u00a0damage caused was regarded as the\u00a0greatest in\u00a0history until 1997. This may be illustrated by two contemporary reactions. \u201cMy\u00a0heart bled at the\u00a0sight of\u00a0such a\u00a0disaster,\u201d said the\u00a0provincial governor, Count Heinrich Larisch-M\u00f6nnich, in\u00a0an\u00a0emotional opening address to that year\u2019s\u00a0session of\u00a0the\u00a0Silesian\u00a0parliament in\u00a0Opava. A\u00a0\u201cview from below\u201d was, among others, captured by the\u00a0Silesian\u00a0prose writer Ludmila Ho\u0159k\u00e1 (1892\u20131966) in\u00a0her short story Velk\u00e1 voda<\/span><\/em>\u00a0[<\/span>The\u00a0Great Flood], which depicts the\u00a0coexistence of\u00a0villagers with the\u00a0river: \u201cAnd Franti\u0161ka still remembers the\u00a0great flood of\u00a01903, when boats passed through our village (near Krava\u0159e) as if it were Venice, when water flowed into cottages through the\u00a0windows, and when two of\u00a0them collapsed as though made of\u00a0gingerbread. At that time even the\u00a0authorities made an\u00a0effort, sending soldiers to help, and firefighters also arrived from somewhere \u2013 but that is useful when there is a\u00a0fire, and since there was no fire anywhere, they at least extinguished their thirst.\u201d<\/span><\/span><\/p>\n
The\u00a01903 flood from a\u00a0hydrological perspective in\u00a0historical and other sources<\/h3>\n
The\u00a0flood of\u00a010\u201311\u00a0July 1903 in\u00a0the\u00a0Odra basin, which until the\u00a0July 1997 flood had been the\u00a0most destructive modern flood event in\u00a0the\u00a0region, represented a\u00a0major natural disaster for this area. It was caused by extremely high two-day precipitation totals. On 9\u00a0July, the\u00a0centre of\u00a0the\u00a0precipitation was located in\u00a0the\u00a0Hrub\u00fd Jesen\u00edk region, before shifting to Moravian-Silesian\u00a0Beskydy on the\u00a0following day.<\/p>\n
Already in\u00a0the\u00a0same year, several authors described the\u00a0situation in\u00a0Jesen\u00edky in\u00a0detail, for example Neu verbesserte Auflage der Hochwasser-Katastrophe am 10. und 11. Juli 1903 in\u00a0politischen Bezirk Freiwaldau (1903)\u00a0[3], possibly\u00a0[4] or Landesanstalt f\u00fcr Gew\u00e4sserkunde (1904)\u00a0[5]. The\u00a0event was subsequently revisited by other authors, including Zeman\u00a0(1961)\u00a0[6], Pol\u00e1ch and G\u00e1ba (1998)\u00a0[7], \u0160tekl\u00a0et\u00a0al. (2001)\u00a0[8], and \u0158ez\u00e1\u010dov\u00e1 et\u00a0al. (2003)\u00a0[9]. The\u00a0120th\u00a0anniversary of\u00a0the\u00a0flood in\u00a0Jesen\u00edky was commemorated by Hal\u00e1sov\u00e1 (2023)\u00a0[10].<\/p>\n
<\/a>\nFig. 2. Photo from the\u00a01903 flood in\u00a0the\u00a0area of\u00a0Peka\u0159sk\u00e1 street (source: Provincial Archives of\u00a0Opava, archival number cz227205010\/\/1062\/\/1\/2\/1\/1\/2\/\/23+43)<\/h6>\n
In\u00a0the\u00a0B\u011bl\u00e1 basin\u00a0and in\u00a0the\u00a0catchments of\u00a0streams flowing into K\u0142odzko Nysa, flood marks comparable to those in\u00a0Opava are not available. An\u00a0exception is the\u00a0statue of\u00a0St Florian\u00a0in\u00a0Vidnava, on which flood marks from 1903, 1997, and 2024 can\u00a0be found. The\u00a0statue is on the\u00a0right bank approximately 70\u00a0m from the\u00a0watercourse and about 300\u00a0m from Vidnava hydrological station (Vidn\u00e1vka), which is on the\u00a0left bank further downstream. It is known that Vidn\u00e1vka reached peak water levels of\u00a0370\u00a0cm on 7\u00a0July 1997 and 453\u00a0cm on 15\u00a0September 2024. The\u00a0flood mark from 1903 is situated between them, indicating that Vidn\u00e1vka most probably exceeded 400\u00a0cm. At that time, however, the\u00a0river had not yet undergone significant channel regulation, which introduces additional uncertainty into the\u00a0estimates. The\u00a0situation is even more complicated on the\u00a0B\u011bl\u00e1, where flood marks documenting individual events are absent. According to the\u00a0measured data, the\u00a0B\u011bl\u00e1 at Mikulovice reached a\u00a0peak water level of\u00a0475\u00a0cm on 15\u00a0September 2024 and 407\u00a0cm on 7\u00a0July 1997. For 1903, no measured data are available and it is therefore necessary to rely solely on descriptions of\u00a0the\u00a0event in\u00a0historical sources. Zeman\u00a0(1961)\u00a0[6] states that a\u00a0water level of\u00a0598\u00a0cm was recorded in\u00a0\u010cesk\u00e1 Ves at the\u00a0toll bridge. Other sources refer to a\u00a0level of\u00a0just under 6\u00a0metres. The\u00a0present-day hydrological station and the\u00a0bridge are 8\u00a0km apart in\u00a0a\u00a0straight line and approximately 11\u00a0km apart along the\u00a0river channel. The\u00a0water level measured at the\u00a0bridge may have been influenced by the\u00a0bridge itself \u2013 specifically by its blockage and the\u00a0resulting backwater effect. Furthermore, it is an\u00a0indisputable fact that one of\u00a0the\u00a0narrowest sections of\u00a0the\u00a0B\u011bl\u00e1 river is located approximately 700\u00a0m upstream, which could have significantly affected the\u00a0water level at the\u00a0bridge. It is also documented that a\u00a0breach occurred in\u00a0the\u00a0regulation embankment at Sta\u0159\u00ed\u010d, i.e. upstream of\u00a0the\u00a0bridge. In\u00a0view of\u00a0the\u00a0above findings and the\u00a0enormous material damage, it may be assumed that the\u00a0B\u011bl\u00e1 river in\u00a0Mikulovice most probably reached a\u00a0peak water level exceeding 400\u00a0cm in\u00a01903. Whether the\u00a0level was even higher would require further and more detailed analysis.<\/p>\n
If attention is focused on the\u00a0town of\u00a0Opava itself, the\u00a0study by K\u0159\u00ed\u017e, Sochorec, and K\u0159\u00ed\u017e (1963)\u00a0[11] probably provides the\u00a0most detailed analysis of\u00a0the\u00a01903\u00a0flood. The\u00a0authors focused primarily on hydrological and hydraulic evaluation based on the\u00a0preserved data available at the\u00a0time. Given the\u00a0period in\u00a0which the\u00a0study was conducted, it is evident that GIS, mathematical models, and digital data could not yet be used.<\/p>\n
Systematic observations of\u00a0water levels on the\u00a0Opava began\u00a0in\u00a01895. Until regulation of\u00a0the\u00a0Opava channel and the\u00a0profile in\u00a01912\u20131913, the\u00a0water level on 11\u00a0July 1903 remained the\u00a0highest on record, reaching 525\u00a0cm. Using the\u00a0rating curve valid at the\u00a0time for the\u00a0surveyed Pil\u0161\u0165sk\u00fd Bridge profile, the\u00a0Hydrological Department in\u00a0Opava calculated a\u00a0discharge of\u00a0454.5\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>.The\u00a0authors of\u00a0study\u00a0[11], however, questioned this value and, using a\u00a0formula for mean\u00a0cross-sectional velocity with values of\u00a01.7\u20131.8\u00a0m\u2009\u2219\u2009s-1<\/sup> and a\u00a0flow area of\u00a0138\u00a0m\u00b2, proposed a\u00a0revised peak discharge of\u00a0235\u2013248\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>. The\u00a0authors themselves nevertheless emphasised that, at the\u00a0time of\u00a0the\u00a0evaluation, they did not have data on flow areas and velocities for the\u00a0left and right floodplains, which are essential parameters for the\u00a0assessment of\u00a0an\u00a0extreme flood event. Likewise, in\u00a0their subsequent calculations, the\u00a0authors revised the\u00a0value of\u00a0the\u00a0mean\u00a0cross-sectional velocity during the\u00a0flood peak from 1.8\u00a0m\u2009\u2219\u2009s-1<\/sup> and, using Manning\u2019s\u00a0equation, proposed revised values in\u00a0the\u00a0range
\nof\u00a02.32\u20133.11\u00a0m\u2009\u2219\u2009s-1<\/sup>. Using these values, they proposed a\u00a0peak discharge in\u00a0the\u00a0range of\u00a0283\u2013404\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>. The\u00a0mean\u00a0value of\u00a0343.5\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> more or less corresponds to the\u00a0peak discharge value of\u00a0360\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> reported by Brosch\u00a0[2]. For\u00a0the\u00a01903\u00a0flood, Brosch\u00a0[2] reports peak discharges of\u00a0750\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> for Ostravice\u00a0\/ Moravsk\u00e1 Ostrava profile, 400\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> for Odra \/ Svinov profile, and 1,500\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> for Odra \/ Bohum\u00edn profile, where sources describe the\u00a0event as the\u00a0greatest flood disaster in\u00a0the\u00a0history of\u00a0the\u00a0town\u00a0[4]. In\u00a0discussing the\u00a0missing data, the\u00a0authors also point out the\u00a0absence of\u00a0information on water-surface slope within\u00a0the\u00a0channel and floodplain\u00a0areas, which can\u00a0nowadays be successfully addressed using 1D and 2D hydraulic modelling. Study\u00a0[8] also employed an\u00a0approach based on the\u00a0antecedent precipitation index (API), in\u00a0which the\u00a0maximum possible peak discharge derived using this method was estimated at 290\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>; however, from the\u00a0perspective of\u00a0simulations of\u00a0this episode using rainfall\u2013runoff models with an\u00a0adequate API, this value appears to be substantially underestimated. Likewise, it does not correspond to the\u00a0differences in\u00a0water levels indicated by the\u00a0preserved flood marks in\u00a0Opava. To illustrate the\u00a0wide range of\u00a0estimated values, reference may also be made to the\u00a0conclusions of\u00a0the\u00a0team led by Br\u00e1zdil and Kirchner (2007)\u00a0[1], as well as the\u00a0earlier monograph by Br\u00e1zdil et al.\u00a0[12], in\u00a0which the\u00a0authors interpret the\u00a0flood peak as corresponding approximately to a\u00a0Q50 event. Such a\u00a0return period, however, would not have caused the\u00a0catastrophic damage reported in\u00a0contemporary sources for the\u00a0area between Opava and D\u011bhylov.<\/p>\nThe overall situation in the town of Opava is further illustrated by reports in the contemporary regional press, for example the newspaper Grenzbote des nordwestlichen M\u00e4hrens [13] (translated from German): \u201cThe situation in Opava was also extremely dangerous. First, Kate\u0159inky (Katharein) was flooded, after which\u00a0the local fire brigade and volunteer firefighters from Opava intervened. An hour later, the floodwaters also inundated Ratibo\u0159 suburb (Ratiborer Vorstadt). A company of the territorial defence force was called in and, together with firefighters, began evacuating residents from houses situated in lower-lying areas by the light of torches and lanterns. On Friday night (10 July), the water level in the Opava had already risen to the bridge, and it was expected at any moment that the river would overflow its banks. At that moment, two additional companies of imperial infantry arrived at the bridge. A third company was dispatched to the sugar refinery in Kate\u0159inky, where the situation had become critical. Water had entered many houses and in some places even covered the ground-floor windows, making the rapid evacuation of endangered residents essential. Many refused to leave their homes and had to be evacuated by force.\u201d The newspaper Deutsches Volksblatt f\u00fcr M\u00e4hren und Schlesien (15 July 1903, No. 56, Vol. 27, p. 4) reported: \u201cThe news from Opava is distressing. It was reported from there on the 12th of this month: Although the water has receded considerably, Ratibo\u0159 suburb and the neighbouring municipality of Kate\u0159inky still resemble one vast lake, from which half-collapsed houses, tree trunks, and pieces of furniture protrude. With the assistance of an engineer company that had arrived late the previous evening from Krak\u00f3w, it proved possible at dawn today to reach those parts of Ratibo\u0159 suburb that had until then been completely cut off from the outside world. Among them was Schwarzgasse (Black street), which presented a terrifying scene of suffering. The day before yesterday, water had entered the ground-floor dwellings there with such speed that the inhabitants barely escaped with their lives. Meanwhile, the military authorities organised supply convoys which transported bread, rolls, and milk by pontoon to houses affected by the flood. The first convoy was led by the provincial president, Count Thun-Hohenstein, who also personally participated in the distribution of food, at times standing chest-deep in water. During the course of the day, thirteen houses collapsed in Ratibo\u0159 suburb after their foundation pillars had been undermined by water, and many more houses were on the verge of collapse. A man whose identity has not yet been established was killed beneath the ruins of one of the collapsed houses. The greatest disaster currently causing concern is the shortage of water resulting from the flood, which has affected Opava since this morning. As already reported, the waterworks was forced to suspend operations yesterday as a result of the flooding, because its water supplies had been contaminated by infiltrating groundwater. Thus, since this morning the town has effectively been without water. The municipal park and the military shooting range are flooded, the waterworks is under water, and operations had to be suspended. In Kate\u0159inky and Ratibo\u0159 suburb, the flood caused extensive damage. Cesspits in the houses were undermined and lifted by the water. N\u00e1kladn\u00ed street (Lastenstrasse) and all the surrounding houses and gardens are completely flooded, while in Peka\u0159sk\u00e1 street (B\u00e4ckergasse) the water reached as far as 400 paces from Horn\u00ed n\u00e1m\u011bst\u00ed (Oberring). In Parkstra\u00dfe (the area of present-day Sadov\u00e1 street), one house collapsed. The inhabitants were rescued in time. Many shops and offices had to operate with a greatly reduced number of employees because staff were unable to reach the town. For this reason, and also because the printing works was under water, the newspaper Freie schles. Presse could not be published.\u201d Lidov\u00e9 noviny then provided more general information (12 July 1903, No. 157, Vol. 11): \u201cFrom Opava: Kate\u0159inky was flooded yesterday. Two additional companies of troops were\u00a0yesterday dispatched to Krnov, which is under water and completely cut off from all railway connections, in\u00a0order to provide assistance\u2026 Yesterday at 4 p.m., railway traffic between Opava and Krnov had to be suspended because the\u00a0continually rising water had severely damaged the\u00a0line. At 5 p.m., an\u00a0attempt was made to dispatch a\u00a0relief train\u00a0carrying three battalions of\u00a0troops to Krnov, but the\u00a0train\u00a0was forced to stop at \u00davalno station, from where the\u00a0soldiers had to continue on foot to Krnov. The\u00a0military command in\u00a0Opava requested the\u00a0immediate deployment of\u00a0engineering corps units from Krak\u00f3w to the\u00a0endangered districts. According to incoming reports, the\u00a0flood had reached dimensions such as had never before been experienced in\u00a0Silesia. The\u00a0damage is enormous\u2026 14\u00a0July 1903, No. 158, Vol. 159 \u2026 All municipalities along the\u00a0Northern Railway line from Opava to Svinov are flooded.\u201d<\/p>\n
Interesting sources of information are also provided by the recollections of eyewitnesses who were expelled to Germany after the Second World War. Two contributions dealing directly with the 1903 flood in Opava were published in Troppauer Heimat-Chronik in 1953 and 1963. According to Franz [14], the water level of the Opava river at the \u201clarge bridge\u201d was approximately 1.7 m above normal level. This level was recorded at noon on Friday 10 July. The author lived directly at Ratibo\u0159sk\u00e1 street No. 40 \/Ratiborer Stra\u00dfe 40\/ in Opava and, later that same afternoon \/10 July\/, the water had risen to the height of the shed roof. \u201cThe water, however, was filthy and smelled terrible because it was mixed with the contents of cesspits and manure heaps.\u201d From the afternoon onwards, the family remained in the attic of the house. Crossing the bridge had already become impossible, since water was flowing across it like a wild river. Nevertheless, on the following morning \/11 July\/, firefighters were still working on the bridge, attempting to remove beams trapped in the bridge structure in order to clear the passage for the water. The water level inside the house reached a height of 167 cm \/on the morning of 11 July\/. \u201cWe spent the night before Sunday, 12 July 1903, in relative peace, because the rain had stopped and the water had begun to recede slowly. On Sunday, the municipal building authority and volunteers erected temporary footbridges in the streets. By Sunday evening, the water had receded sufficiently for our\u00a0house to be free of\u00a0flooding. Every room was covered in\u00a0mud up to waist height.\u201d Franz further notes that the\u00a0horrors of\u00a0the\u00a0floods, which frequently occurred throughout the\u00a0history of\u00a0Opava, are still recalled by the\u00a0unofficial name of\u00a0Ratibo\u0159 suburb: \u201cNasses Viertel<\/em>\u201d (\u201cWet Quarter\u201d). The\u00a0second eyewitness, Otto Schreiber, lived nearby at Ratibo\u0159sk\u00e1 No. 30 \/Ratiborer Stra\u00dfe 30\/. Schreiber\u00a0[15] describes how the\u00a0water level rose to just 5\u00a0cm below the\u00a0windows of\u00a0his room and how his family had to live for several days on the\u00a0first floor of\u00a0a\u00a0neighbour\u2019s\u00a0house. He further notes that this flood, followed by the\u00a0flood of\u00a0September 1910, subsequently led to extensive regulation of\u00a0the\u00a0Opava river channel.<\/p>\n
Historical plans, sketches, and maps constitute an\u00a0important source of\u00a0information, particularly regarding the\u00a0topography of\u00a0the\u00a0study area. Fig.\u00a03<\/em> shows a\u00a0site plan\u00a0of\u00a0part of\u00a0the\u00a0town of\u00a0Opava, focusing on Ratibo\u0159 suburb and Kate\u0159inky, dating from 1826 (Provincial Archives in\u00a0Opava). In\u00a0comparison with the\u00a0Imperial Imprints of\u00a0the\u00a0Stable Cadastre (Moravia and Silesia were surveyed between 1826 and 1836), differences are apparent in\u00a0the\u00a0representation of\u00a0the\u00a0built-up area of\u00a0Kate\u0159inky at that time, see Fig.\u00a04<\/em>. Differences in\u00a0the\u00a0depiction of\u00a0buildings are also visible in\u00a0the\u00a0Second Military Survey (Moravia and Silesia were surveyed between 1836 and 1840). It is also necessary to take into account the\u00a0fact that Opava was affected by one of\u00a0the\u00a0more significant historical floods precisely in\u00a01826, after which one of\u00a0the\u00a0historical regulations of\u00a0the\u00a0Opava channel was conducted; further floods subsequently occurred in\u00a01829, 1831, and 1838\u00a0[1, 2].<\/p>\n
<\/a>\nFig. 3. Site plan\u00a0of\u00a0Kate\u0159inky and Racib\u00f3rz suburb from 1826 (source: Provincial Archives of\u00a0Opava, archival number 688, inv. no. 50, signature 37\/2)<\/h6>\n
<\/a>\nFig. 4. Extract from\u00a0the\u00a0Imperial Imprint of\u00a0the\u00a0Stable Cadastre for Kate\u0159inky and the\u00a0Racib\u00f3rz suburb (source: \u010c\u00daZK)<\/h6>\n
Data and methods for hydrological reconstruction<\/h3>\n
As already mentioned in the introduction, the main tools used for the reconstruction of the flood were GIS software (ESRI ArcMap, ArcHydro, SAGA GIS, GRASS GIS), rainfall\u2013runoff models (HEC-HMS), and hydrodynamic models (HEC-RAS, MIKE 11\/21c). The Imperial Imprints of the Stable Cadastre and the Second Military Survey were the principal sources of information on historical topography and, to some extent, elevation data. Among the most important historical cartographic materials provided by the Provincial Archives in Opava\u00a0were the Site Plan of the Town of Opava from 1826, the Plan of the Drainage Situation of the Sugar Refinery in Opava from 1854, and in particular the Plan der Landeshauptstadt Troppau mit der neuen Wasserleitung und den Stadterweiterungsgr\u00fcnden by Eduard Labitzke from 1876. After transformation into the S-JTSK coordinate system using control points, these materials were used for the digitisation of the drainage network, specifically the Opava channel itself and the historical millraces. Following digitisation in GIS, these features were converted into the schematic formats required for the DHI MIKE and HEC-RAS hydrodynamic models. The digital terrain model was subsequently incised along the historical course of the Opava channel using the ArcHydro toolset for the ESRI platform and the r.carve module in GRASS GIS. The historical millraces were likewise incised into the terrain model; however, readily usable geodetic data describing the channel parameters and diversion structures of these features are not available (parameters of the channels can also be derived from historical water registers; however, this represents a highly demanding\u00a0task in\u00a0terms of\u00a0both archival research and time, which is one of\u00a0the\u00a0reasons why this article emphasises the\u00a0much more efficient possibilities offered by GIS tools). Nevertheless, during such an\u00a0extreme flood event, the\u00a0influence of\u00a0these hydraulic structures on the\u00a0extent of\u00a0the\u00a0inundated area was probably only marginal. Moreover, most of\u00a0the\u00a0millraces were situated in\u00a0the\u00a0right-bank zone, whereas the\u00a0most severely affected area, Kate\u0159inky, lies in\u00a0the\u00a0left-bank zone. The\u00a0only exception was a\u00a0millrace that approximately followed the\u00a0route of\u00a0the\u00a0present-day streets Partyz\u00e1nsk\u00e1, Holasick\u00e1, and Na Pot\u016f\u010dku.<\/p>\n
Sensitivity analysis using the\u00a0Markov Chain\u00a0Monte Carlo method was conducted primarily within\u00a0the\u00a0HEC-HMS rainfall\u2013runoff model, specifically for the\u00a0possible ranges of\u00a0infiltration values, Curve Number (CN) values, and concentration times, which depend both on catchment morphology and land use. In\u00a01903, forest cover in\u00a0the\u00a0Opava basin\u00a0was undoubtedly less extensive than\u00a0today; therefore, CN values for the\u00a0individual sub-catchments were set within\u00a0the\u00a0range from 65 (forest, HSG A) to 92 (arable land, HSG D), with 250\u00a0samples generated for the\u00a0specified interval range. Because CN values are primarily derived for agricultural soils, values for forest soils in\u00a0the\u00a0upper parts of\u00a0the\u00a0catchment were derived using nomograms for forested areas; see, for example, Haan\u00a0et al.\u00a0[16] and Mishra and Singh\u00a0[17]. The\u00a0AMC (Antecedent Moisture Conditions) parameter was not modified because both precipitation totals and the\u00a0antecedent precipitation index were known. At the\u00a0same time, conversion equations between the\u00a0SCS-CN method and the\u00a0Green\u2013Ampt method (see, for example, Mishra and Singh\u00a0[17] and Bedient et al.\u00a0[18]) were used to define the\u00a0interval range for infiltration velocity.<\/p>\n
The\u00a0results of\u00a0the\u00a0HEC-HMS simulations with Markov Chain\u00a0Monte Carlo sensitivity analysis are shown in\u00a0Fig.\u00a05. For the\u00a0present-day Opava gauging profile, the\u00a0peak discharge obtained using the\u00a0SCS-CN method (and similarly for the\u00a0Green\u2013Ampt method) ranged from 389\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> to 529\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>, with a\u00a0mean\u00a0peak discharge of\u00a0444\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>. This value corresponds relatively well to the\u00a0peak discharge derived from the\u00a0historical rating curve (454\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>).<\/p>\n
<\/a>\nFig. 5. Comparison of\u00a0the\u00a0simulated flood peak of\u00a0the\u00a01903 flood using the\u00a0HEC-HMS\u00a0model for the\u00a0Opava profile with uncertainty analysis using\u00a0the\u00a0Markov\u00a0chain\u00a0Monte Carlo method<\/h6>\n
The\u00a0simulations performed in\u00a0the\u00a0HEC-HMS rainfall\u2013runoff model were subsequently followed by simulations in\u00a0the\u00a0MIKE and HEC-RAS hydraulic models in\u00a0both 1D (reconstructed course and condition of\u00a0the\u00a0channel according to preserved historical sources and maps) and 2D (state of\u00a0the\u00a0floodplain\u00a0areas based on preserved maps and plans). Flood-mark elevations (Ratibo\u0159sk\u00e1 street and the\u00a0industrial area on Sadov\u00e1 street) and the\u00a0extent of\u00a0inundated areas in\u00a0comparison with the\u00a0Q100<\/span> flood and the\u00a0floods of\u00a01997 and 2024 were used for verification (Fig.\u00a06<\/span><\/em>). In\u00a0terms of\u00a0the\u00a0flood-mark elevation on Ratibo\u0159sk\u00e1 street, the\u00a0simulated water level corresponding to a\u00a0discharge of\u00a0438\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> provided the\u00a0best agreement (Fig.\u00a07<\/span><\/em>). If historical sources describing the\u00a0situation in\u00a0Krava\u0159e and other settlements further downstream are also taken into account, this value may be regarded as realistic, including in\u00a0comparison with the\u00a0evaluated peak discharge of\u00a0670\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> for the\u00a0September 2024 flood (the\u00a0difference between the\u00a0flood-mark elevation and the\u00a009\/2024 flood level is 23\u00a0cm).<\/span><\/p>\n
<\/a>\nFig. 6. Simulated water levels and flood extent in 1D and 2D hydraulic models (HEC-RAS and DHI MIKE) for the peak discharge of 438 m3\u2009<\/sup>\u2219 s-1<\/sup>. The red line is the Q100<\/sub> boundary (388 m3\u2009<\/sup>\u2219 s-1<\/sup>) for the present-day riverbed and floodplain. The base map is Labitzky\u2018s 1876 plan of Opava<\/h6>\n
<\/a>\nFig. 7. Simulated water level in the HEC-RAS model for a peak discharge of 438\u00a0m3<\/sup>\u2009\u2219\u2009s-1<\/sup> in\u00a0the area of \u200b\u200bRatibo\u0159sk\u00e1 street. The elevation of the flood mark on building no.\u00a043 is 249.45 m above sea level (source: own geodetic measurements)<\/h6>\n
DISCUSSION<\/h2>\n
The\u00a0reconstruction of\u00a0the\u00a0July 1903 flood in\u00a0the\u00a0Opava River basin\u00a0is, naturally, subject to a\u00a0range of\u00a0uncertainties arising primarily from the\u00a0limited availability of\u00a0direct hydrological measurements, changes in\u00a0the\u00a0morphology of\u00a0channels and floodplain\u00a0areas, and substantial transformations in\u00a0land use from the\u00a0beginning of\u00a0the\u00a020th century to the\u00a0present day. Nevertheless, based on the\u00a0combination of\u00a0historical sources, flood marks, contemporary testimonies, and later hydrological analyses, it may be concluded that this was an\u00a0exceptionally extreme event whose impacts were comparable to \u2013 and in\u00a0some profiles probably even more severe than\u00a0\u2013 those of\u00a0the\u00a0floods of\u00a01997 and 2024. The\u00a0considerable variability in\u00a0estimated peak discharges for the\u00a0Opava river reported in\u00a0the\u00a0literature illustrates the\u00a0limitations of\u00a0traditional calculation approaches based on simplified hydraulic relationships and incomplete data. In\u00a0particular, neglecting floodplain\u00a0conveyance and uncertainties in\u00a0the\u00a0estimation of\u00a0mean\u00a0cross-sectional velocity and hydraulic water-surface slope may have led to underestimation of\u00a0peak discharges in\u00a0some studies. Estimates of\u00a0flood peaks below the\u00a0Q50<\/sub>\u00a0level (currently 312\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup>) are inconsistent both with the\u00a0documented extent of\u00a0damage within\u00a0the\u00a0built-up area of\u00a0Opava and in\u00a0settlements further downstream as far as D\u011bhylov profile, and with the\u00a0preserved flood marks. Likewise, the\u00a0peak discharge value of\u00a0290\u00a0m\u00b3\u2009\u2219\u2009s-1<\/sup> reported by K\u0159\u00ed\u017e et al.\u00a0[11] as the\u00a0minimum estimate within\u00a0their range of\u00a0values may also be regarded as underestimated. Modern hydrodynamic modelling tools, which make it possible to explicitly incorporate the\u00a0morphological and hydraulic parameters of\u00a0floodplain\u00a0areas and the\u00a0numerical solution of\u00a0flow in\u00a01D and 2D, represent a\u00a0substantial advance in\u00a0this respect and offer the\u00a0possibility of\u00a0refining historical estimates. Conversely, the\u00a0combination of\u00a0extreme precipitation totals, high antecedent catchment saturation, and the\u00a0historical state of\u00a0the\u00a0river network indicates an\u00a0event approaching extremes with a\u00a0very low probability of\u00a0occurrence.<\/p>\n
An\u00a0interesting aspect is the\u00a0comparison of\u00a0the\u00a01903 flood with the\u00a0floods of\u00a01997 and 2024. Although the\u00a0present-day landscape differs substantially from that of\u00a0the\u00a0Opava basin\u00a0at the\u00a0beginning of\u00a0the\u00a020th century \u2013 among other things in\u00a0terms of\u00a0forest cover, the\u00a0higher proportion of\u00a0built-up areas and technical river regulation works, and the\u00a0presence of\u00a0water reservoirs on Moravice\u00a0\u2013 the\u00a0influence of\u00a0these historical differences gradually diminishes once the\u00a0infiltration or retention capacity of\u00a0the\u00a0landscape is exceeded by extreme precipitation, and surface runoff becomes dominant. Under such conditions, it is more appropriate to focus on the\u00a0fluvial component of\u00a0flood dynamics and on the\u00a0historical state of\u00a0river channels, the\u00a0influence of\u00a0millraces and other historical hydraulic structures, and especially the\u00a0morphology and transformation of\u00a0built-up areas within\u00a0floodplains, which may represent an\u00a0interesting subject for further research. These factors also support the\u00a0initial hypothesis that, during such extreme events, the\u00a0dominant role is played primarily by meteorological forcing together with the\u00a0morphometric and hydrological characteristics of\u00a0the\u00a0catchment.<\/p>\n
The\u00a0results of\u00a0this discussion also underline the\u00a0importance of\u00a0historical floods for present-day water management planning. The\u00a01903 flood played an\u00a0important role in\u00a0historical considerations of\u00a0flood-protection measures, and its more detailed quantitative reconstruction may, among other things, contribute to a\u00a0more objective assessment of\u00a0the\u00a0design parameters of\u00a0water-management structures and measures, such as Nov\u00e9 He\u0159minovy reservoir, as well as to a\u00a0better understanding of\u00a0the\u00a0actual flood risk in\u00a0the\u00a0Opava River basin.<\/p>\n
CONCLUSION<\/h2>\n
The flood of July 1903 in the Odra and Opava basins undoubtedly represents an extreme hydrological event whose quantitative description is subject to considerable uncertainty arising from the absence of direct discharge measurements, incomplete information on hydraulic conditions, and substantial changes in the morphology of river channels and floodplain areas since the beginning of the 20th century. Nevertheless, the analysis of historical sources in combination with the results of rainfall\u2013runoff and hydrodynamic modelling makes it possible to determine a realistic range of potential peak discharges and the course of the flood wave. The results of the rainfall\u2013run-off simulations confirm the\u00a0key role of\u00a0extreme two-day precipitation totals and the\u00a0degree of\u00a0antecedent catchment saturation; under such conditions, the\u00a0retention capacity of\u00a0soils becomes rapidly saturated and the\u00a0importance of\u00a0differences in\u00a0land use decreases. At the\u00a0same time, the\u00a0modelling results show that approaches based exclusively on simple antecedent precipitation indices or simplified empirical relationships frequently lead to errors in\u00a0the\u00a0estimation of\u00a0peak discharges and flood-wave volumes. In\u00a0this context, hydrodynamic modelling in\u00a01D and 2D environments appears to be an\u00a0essential tool for the\u00a0reconstruction of\u00a0historical floods, since it makes it possible to incorporate floodplain\u00a0conveyance, the\u00a0spatial variability of\u00a0flow velocities, and the\u00a0influence of\u00a0local channel constrictions, bridge structures, and historical millraces. These factors may have had a\u00a0crucial influence on water levels within\u00a0the\u00a0built-up areas of\u00a0settlements during the\u00a01903 flood (particularly in\u00a0Opava and Krava\u0159e) and help to explain\u00a0the\u00a0discrepancy between some earlier discharge estimates and the\u00a0documented extent of\u00a0inundation. The\u00a0modelling scenarios also indicate that the\u00a01903 flood probably exceeded the\u00a0characteristics of\u00a0commonly considered design events (e.g. Q\u2085\u2080<\/sub>) and approached extremes with a\u00a0very low probability of\u00a0occurrence (generally Q100 and above). These conclusions are consistent both with the\u00a0preserved flood marks and with qualitative descriptions of\u00a0the\u00a0extent of\u00a0damage contained in\u00a0historical sources. Moreover, comparison with the\u00a0floods of\u00a01997 and 2024 shows that, during extreme precipitation events, catchments exhibit a\u00a0typologically similar response regardless of\u00a0differences in\u00a0present-day land use (through the\u00a0progressive increase in\u00a0the\u00a0proportion, ultimately leading to the\u00a0complete dominance, of\u00a0surface runoff within\u00a0both direct and total catchment runoff), which represents another argument for the\u00a0importance of\u00a0hydrological modelling focused on extreme scenarios. In\u00a0conclusion, it may be stated that the\u00a0combination of\u00a0historical data with GIS-based spatial analyses together with rainfall\u2013runoff and hydrodynamic models represents an\u00a0effective approach to improving the\u00a0reconstruction of\u00a0historical floods and to achieving a\u00a0better understanding of\u00a0their temporal and spatial characteristics. These findings are of\u00a0direct relevance for present-day water management planning, particularly for the\u00a0determination of\u00a0design discharges, the\u00a0calibration of\u00a0extreme scenarios, and the\u00a0professional assessment of\u00a0the\u00a0effectiveness of\u00a0flood-protection measures in\u00a0the\u00a0Opava River basin.<\/p>\n
Acknowledgements<\/h3>\n
The\u00a0authors would like to acknowledge the\u00a0support of\u00a0the\u00a0NAZV project No.\u00a0QL24010054, Impacts of\u00a0Climate Change on Small Forest Catchments and Possibilities for Their Mitigation through Forest Management and Water Management Measures, and the\u00a0project Prediction, Assessment and Sensitivity Analysis of\u00a0Selected Systems, the\u00a0Effects of\u00a0Drought and Climate Change in\u00a0Czechia (PERUN) (No. SS02030040), without whose support the\u00a0preparation of\u00a0this article would have been considerably more time-consuming and complicated. The\u00a0authors also wish to acknowledge the\u00a0support of\u00a0the\u00a0Czech Hydrometeorological Institute\u2019s\u00a0DKRVO programme (Long-Term Concept for the\u00a0Development of\u00a0the\u00a0Research Organisation for the\u00a0Period 2023\u20132027), particularly research areas 6 and 12.<\/em><\/p>\n
The\u00a0Czech version of\u00a0this article was peer-reviewed, the\u00a0English version was translated from the\u00a0Czech original by Environmental Translation Ltd.<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"
The 1903 flood was undoubtedly one of the most significant floods of the 20th century in Moravia and Silesia. Although systematic observa-tions of water levels had already begun at many gauging sites during this period, it is difficult to convert historical water levels recorded at limnigraph stations into present-day equivalents due to historical territorial changes in terms of politics (Prussia versus the present-day Czech Republic), landscape structure (forest cover in the historical and present-day landscape), morphology (terrain and urban development in the affected areas), and water management conditions (the course of the Opava river channel and the condition of water flumes in 1903 and today). Useful (though not always entirely reliable) sources include historical flood marks, photographic documentation, historical maps and plans, reports in the contemporary press, family chronicles, and the recollections of millers, sawyers, and other craftsmen using water power. These sources form a rather heterogeneous body of evidence, and it is therefore necessary to find ways to verify and combine them. One possible approach is to use these data in GIS-based spatial analyses and subsequently as inputs for rainfall\u2013runoff and hydraulic model-ling. Since the team of the Czech Hydrometeorological Institute (CHMI), together with its partners, had already conducted similar analyses (e.g. during the reconstruction of the 1872 historical flood on Bl\u0161anka), they also attempted to apply this approach to the 1903 flood on the Opava river. The results, including a partial uncertainty analysis, are presented in this article.<\/p>\n","protected":false},"author":8,"featured_media":39105,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[94,88,93],"tags":[4150,4151,4121,4152],"coauthors":[3183,3481,4125,4126,4127,4128,4129,4130,3482,4131,4132,4133],"class_list":["post-39107","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-current-issue","category-informatics-cartography-in-water-management","category-two-articles","tag-1903-flood","tag-hydrological-models","tag-opava","tag-reconstruction-of-historical-floods"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/39107","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=39107"}],"version-history":[{"count":6,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/39107\/revisions"}],"predecessor-version":[{"id":39196,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/39107\/revisions\/39196"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media\/39105"}],"wp:attachment":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media?parent=39107"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/categories?post=39107"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/tags?post=39107"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/coauthors?post=39107"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
![\"\"](\"https:\/\/www.vtei.cz\/wp-content\/uploads\/2026\/06\/Unucka-obr-1-1.jpg\")
<\/a>\n