Vaclav Kosacky, Povodi Moravy

Antonin Tuma, Povodi Moravy

Pavel Biza, Povodi Moravy

1. Basic characteristics of the basin

The Morava River in the Czech Republic is one of the most important tributaries of the Danube River. The 22,000 km² Morava River Basin covers 26 percent of the Czech Republic and has a population of 2.7 million. The biggest town in the basin is Brno with about 400,000 inhabitants. The average annual precipitation in the Morava River Basin is 635 mm and the mean annual runoff is appraised at 3,430 million m³.

The Morava River Basin consists of two individual hydrological parts, which in many aspects behave quite differently: the first is the Morava River upstream its confluence with the Dyje River and the second is represented by the Dyje River Subbasin.

The Morava River Basin is of oblong shape. The main tributary to the Morava River is from the left side and it is called the Becva River. The headwaters of both the rivers are characterised by the presence of many brook-size tributaries. They are spread out in the mountainous area of the Jeseniky Mountains (the Morava River), where the average yearly precipitation is 1,400 mm and the Beskydy Mountains (the Becva River), with 1,200 mm of the annual precipitation. Leaving the hilly region, the Morava River has a lowland character, which is typical for the major part of the river with small slope, and an alluvial floodplain developed on both sides.

The Morava River Basin and the Dyje River Basin are dissimilarly equipped with artificial retention reservoirs. In the Dyje River Basin the total volume of flood mitigation and surcharge storage is 43.5 - 85.3 mil. m³, whereas in the Morava River Basin it is only 6.07 mil. m³ though the runoff from this basin is by 30 % larger. The flood retention and mitigation capacity of reservoirs in the Dyje River Basin represents approximately 3 – 6 % of the average annual runoff from the watershed while the capacity of reservoirs in the Morava River Basin is only 0.3 %.

2. THE FLOOD IN JULY 1997

In early July 1997, the weather was characterized by the shift of a large depression from Britain, Scandinavia and southern France to the Czech Republic, and movements of the surrounding cold and the occluded fronts. Much of the subsequent damage and losses were caused by abundant rain in the Morava River spring area on July 4. In the four subsequent days, storm rainfalls occurred not only in the above area, but also in the eastern, central and north-westerly areas of the Morava River drainage area. The most widespread flooding began on 5 July 1997, and the first substantial flush of floodwater came on 7 and 8 July. The situation resulted in the worst flooding in Moravia in all its history.

Due to the situation existing in July of that year, a substantial part of the rainwater was leaving that area immediately and the run-off coefficient was very high during the critical days (0.9 or even higher).

Of the 3,957 km managed by the Povodi Moravy, almost two thousand km (1,954) were affected by the flooding.

Long stretches of water courses were destroyed, with parts of banks being swept away and the growths along the banks and water course branching damaged. Several parallel river beds were formed and protection dikes were damaged. In places, high flow rates caused extensive damage to the valley profiles. The above damage was caused along the rivers Branna, Krupa, Desna, parts of Metra, Morava (from Hanusovice town down to lower Morava river), Roznovska Becva and Bystricka rivers. In addition to river bed damage and destruction of buildings, the floods destroyed parts of roads, bridges, railroad tracts and pipeline and sewerage networks.

In the central reaches of the rivers, and of the Becva and the Morava rivers in particular, the dikes were damaged, parts of river banks swept away, and new deposits were made. The biggest changes to the course and shape of the river bed were found in the case of the Becva river. Overflowing water devastated towns and villages, roads and network system.

In the lower sections of the drainage area, the dikes were eroded and damaged. The overflowing water penetrated to areas of uncontrolled inundation where it behaved in an unpredictable manner and stayed there for different periods of time. Its spread was further influenced by local railroad and road beds and dikes. Inundation areas were up to 5 km wide (13 km in extreme cases) and up to 2.5 m deep.

The damage to the rivers managed by the Povodi Moravy and to water management structures along those rivers totalled 1.7 billion CZK, which includes mainly:

226 km of scoured banks

136 km of damaged protective dikes

91 km of damaged river bed lining

about 280,000 m³ of deposits in river beds

damage to 132 river stabilization structures

Work to repair the damage and reconstruction work on flood protection systems began as soon as local situation permitted it. In the first stage, the focus was on the reconstruction of existing, and on building new, access roads, the removal of barriers (accumulated trees, rubble from destroyed bridges and houses, washings from fields, etc.) and deposits from river beds to allow free passage of water there, and on the repair of the damaged dike systems.

Of the total 251 million CZK invested in repair work carried out in the Morava drainage area in 1997, 130 million CZK came from a government subsidy, 121 million from internal resources of Povodi Moravy, of which 105 million CZK was a loan. The total number of reconstruction projects was 105, with the most extensive and technologically demanding ones being the repair of the Morava embankment at Tlumacov town, the Kyjovka's relief channel embankment and the Morava's Nesyt embankment near Hodonin town.

The amount of flood repair work in 1998 was totalled 444 million CZK and in 1999 totalled 580 million CZK. The objective is to complete all the repairs by the end of the year 2000.

In some places, river beds were changed considerably, and the environmentalists believe that those changes should be used to benefit the environment. The changes in river beds and floodplains are considered as the basis for the river's natural revitalization. After several studies have been made, a number of reaches of the Becva river totalling about 1 to 2,5 km were selected and these will retain its natural course. A more detailed study for the reaches in question is being prepared.

The enormous flooding in July 1997 also triggered discussion on the need for a flood protection system.

In co-operation with Aquatis, the Povodi Moravy presented its "Master plan of flood prevention measures in the River Morava drainage area" to the Ministry of Agriculture of the Czech Republic in May 1998. The material draws on experience from the recent flooding and concludes that a significant part of damage was concentrated in zones and parts of the drainage area characterized by inadequate potential capacity of their water works, namely storage reservoirs and other elements of water management systems capable of storing, diverting or carrying necessary amounts of water. The fact that the protective storage capacity of such reservoirs may play an important regulatory role, and that such reservoirs are able to protect downstream areas against floods became particularly obvious during the floods.

The fact was clearly proven during the catastrophic flooding in July 1997. A preliminary comparison of the situation made at the time of the flooding in different parts of the drainage area showed the positive role of existing storage reservoirs and storm water retention tanks and emphasized the positive role of keeping a part of the discharge in inundation areas. It was also clearly demonstrated how important it is to control the filling and drainage of such inundation areas.

On the basis of this experience, the master plan proposed flood control measures for the affected parts of the Morava drainage area. In the preparation of the master plan, the following principles were applied:

• flood control measures will be designed to protect only settled areas or sites of special public interest; agricultural areas will not be protected;
• measures will be designed for the biggest flood observed, or for Q₁₀₀
• retention abilities of the countryside (natural inundation areas, man-made storm water retention facilities) will be made the maximum use of ;
• the new measures will be linked to the flood control systems already in place to complement and extend them.

Measures are being proposed in several groups of equal importance: new storm water retention tanks, flood release basins, river bed modifications to increase their discharge capacity, building new dikes along rivers and dikes along residential areas. The measures proposed thus represent a synthesis of a number of previously prepared local-scale documents and plans adjusted on the basis of experience gained in the 1997 floods.

The second group of ideas for solving flood protection of Morava river basin includes better land use and increasing water retention in the river basin. The main goal of these ideas is to ensure the maximum effect of water retention along with the minimum investment into building of technical measures in river basin. One sub-section covers local embankments of cities.

The last category of proposals for improving flood protection in the Morava basin is connected to transportation. The idea is to take into account the possible future navigation channel Danube – Oder – Elbe. The ideas were based on utilisation of parts of future navigation channel and its embankments for retention and drain of a flood wave.

3. MATHEMATICAL FLOOD MODEL of Morava and Becva riverS

There is necessary to select from these measures the best possible combination which will be the most suitable regarding flood protection and it was a task for the model use.

The project of creating the flood model of Morava and Becva rivers named "Flood Management in the Czech Republic" started in May 1998 in cooperation with Danish Hydraulic Institute (DHI). This project was carried out under the protection of Ministry of Agriculture of Czech Republic. The following Czech organisations were involved: The Povodi Moravy, the other Czech Water Boards, the Czech Hydrometeorological Institute and the Water Research Institute. On the project was working also company Hydroinform – the agent of DHI in Czech Republic. Project was agreed for period of 16 month and the Danish Environmental Protection Agency (DEPA) supports the project financially.

The mail goal of this project was to transfer know-how in field of using DHI software for building mathematical flood models. Beside this, the flood model of Morava and Becva rivers was constructed and several alternatives were evaluated as alternative flood protection measures.

The model was create using Mike 11 software, The work started by collection of all data which were necessary for model preparation. There were two main groups of data – survey data and hydrology data.

As the survey data were used measured cross sections on the rivers and on the flood plain. More cross sections was collected from several projects and studies on Morava and Becva rivers but part was new measured after the flood.

Very useful tool was digital elevation model of terrain prepared for Morava river basin and covers area nearly 700 km² of flood plain including junction of Morava and Becva rivers in grid 25 m. This digital elevation model of terrain was created by method of air photography, consists from more than 1,5 million of points and was used for cross sections preparation on the flood plains.

Hydrology data were received from Czech Hydrometeorological Institute. Important were time series of measured values from gauging stations on Morava and Becva rivers and on main tributaries. Daily data from 17 gauging stations in the basin were used for the period 1981 – 1986 and for the flood period 1997. Also rainfall data from 130 rain gauge station in the basin were used for the model completed by temperature data and potential evaporation data. Maximum of data was prepared for the flood situation 1997 because it was the best documented flood in Morava river.

Also other data were wery important for the whole work like detailed land use maps in the basin, big amount of photos from the flood 1997, data on sediments in Morava and Becva river etc.

The whole flood model consists of two parts – rainfall-runoff models in Morava river basin and the hydrodynamic model of Morava and Becva rivers.

The total basin 9700 km² large was divided into 40 subcatchments for creating rainfall-runoff models for each river with the basin larger than 100 km². The runoff on these basins is generated based on rainfall information, temperature and evaporation. Calculated values of discharges from rainfall-runoff models are used as input into the hydrodynamic model. The NAM module of Mike 11 was used as modelling tool for these rainfall-runoff models.

The hydrodynamic model of rivers was prepared based on Mike 11 and describes Morava river between town Hanusovice on the upper part and junction with Dyje river – total length is 255 river km. Becva river is included with the length 61 km from the town Valasske Mezirici. There are also some small parts of tributaries Desna, Mostenka and Drevnice rivers.

The hydrodynamic model is prepared as flood model describes water flowing not only through main river channels, but also through flood plains. Flows in flood plains are described by looped system of branches connected with the main river by link channels and connection branches. Objects describing these objects in reality supplement this system. Objects on main river system are mainly bridges and weirs and there are about 300 similar objects, on flood plain they are crossing linear structures like railways or roads. The hydraulic structures inside the model are described as a combination of weirs and culverts. The actual setup of hydrodynamic model consists of nearly 700 branches and more than 6,500 cross-sections including about 650 weirs and almost 400 culverts describing all existing structures.

According to different proposals the several simulations were prepared for consideration of individual variants of flood protection measures which will be proved by the mathematical model:

Scenario 1. Change of land use in the whole catchment. The idea is to increase areas with the meadows and forest to improve retention in the basin and to increase natural water accumulation in the river basin.. It is proposed to change land use on about 20 % of the whole basin (change about 45 % of agricultural land to meadows and forest). In the rainfall runoff NAM models this changes were simulated by the parameter change for each subbasin.

Scenario 2. A group of proposals from towns and villages along the rivers aim at strengthening local embankments for the protection of these urban areas. This simulation can evaluate the impact of dikes and local embankments on flood discharge in the basin because it is expected that smaller flooded area create higher peaks during flood. In the model this scenario was simulated by the setup change of the model according to these proposals.

Scenario 3. Using water reservoirs for flood protection. On Morava and Becva river there are proposed several places for new reservoirs or dry polders and this simulation can determine the effect of these reservoirs for lower part of rivers. There are three selected locations for retention: Hanusovice and Mohelnice on the Morava river and Teplice on the Becva river.

A reservoir upstream of Hanusovice town can have volume about 30 – 40 millions m³. At Mohelnice town it is proposed to built a dry retention polder with volume of about 27 millions m³. The most promising retention area is at Teplice town where it is possible to construct a reservoir with 24 m high levee and maximum volume of 169 millions m³ or a dry retention polder with an 11 m high levee and maximum capacity of 38 millions m³.

Scenario 4. Using navigation channel for flood protection. In the basin there is a old idea to build a navigation channel Danube – Oder – Elbe and such a channel may be applied as an additional retention storage and safety drain of flood wave.

Impact of scenario 1 is global, it means – the impact is visible along whole modelled area, and this impact is practically very similar for flood Q1997 as for flood about Q_{100 yr,}. The reduction of discharges was calculated in range of 6 – 8 %. Higher reductions are observed for Morava river, mainly in profile Olomouc, just before junction with Becva river. These differences between Morava and Becva rivers are due to different current state of land use on this catchments and different proposals for changes of land use in this scenario. Becva river catchment is already forest and the proposed area for changes from agricultural land to meadows and forests are approximately half of area for changes on Morava river catchment.

Final impact on shape of flood wave is recognisable on approximately 4 hours of delay on start of flood. Descending part of wave is more less the same as for flood 1997.

Impacts of scenario 2 is local, but in contrast to scenario 1, here it causes a local increase of discharges. In upper parts of the rivers, where no significant changes in the flood plains are made, the impact is negligible. Observable effect are seen from profiles Moravicany on Morava river and Teplice on Becva river. Changes in peak discharges varies from 0 to +12 %, and the biggest difference was observed in profile Straznice. This scenario also has a significant impact on the timing of floods. Wave acceleration varied from 0 to 24 hours in comparison with flood 1997. Biggest difference in time of peak between scenarios 0 and 2 were observed in profile Straznice. Realization of this scenario will protect all protected areas, but it will require a reconstruction of 90% of all bridges and weirs in protected areas in addition to protection dikes.

Effect of scenario 3 is possible to see from profiles of polders (Teplice – Becva river and Mohelnice – Morava river) up to end of model. However major effect of polders tails off after junction of both rivers and after this junction (from profile Kromeriz) is this effect expressively smaller. For this scenario is typical chance to utilize different kinds of manipulation rules (longer lag time, start of accumulation of water for different discharges, etc). From results of tested variant imply, that manipulation on polder Teplice comply very well with requests for effective wave transformation. I the case of reservoirs alternative the effect of this scenario is significant along whole Morava and Becva rivers – also after their junction. Below junction of Morava and Becva rivers is recognizable big difference between variants of polder and reservoir.

Effect of scenario 4 is significant in profiles, where are included proposed parts of navigation channel and the reduction is very similar for flood 1997 as for reduced flood. At profiles, where were not designed any changes as a part of scenario 4, the flood pass through faster and with practically unchanged or with little increased culmination discharges.

4. CONCLUSIONS AND NEXT STEPS

The planned outputs of the project were fully achieved. The flood model was created and an accurate description of floods including the extreme 1997 event has been obtained. The model has been applied to the proposed scenarios to illustrate their effect on floods of different magnitude. The model simulations have shown that floods in the Morava catchment may be significantly reduced by implementation of appropriate measures.

A number of questions remain, for example:

• Which measures would be the most efficient lifesavers?
• What are the benefits compared to the costs of these measures?
• Which measures are most attractive economically?

For that reason and because the cooperation between MRBA and DHI was successful a second phase of the project started in January 2000 and it will finish in June 2001.

5. REFERENCES

Ammentorp, H.C., P. Biza, and J. Spatka, "Improving flood management in the Czech Republic", Proceedings of the 3^rd DHI software user conference, 1999.

Kosacky, V, and P. Kutalek, " Damage caused by the 1997 floods, its repair and proposed alternatives of flood control measures ", Proceedings of the NATO Advanced Research Workshop on Coping with Floods: Lessons Learned from Recent Experience, 1999.

Biza, P., and V. Gimun, " Using of Mike 11 flood model of the Morava river in the Czech Republic for evaluating various flood measures proposals ", Proceedings of the 4^th international conference Hydroinformatics 2000, 2000.