Search

Rijeke su složeno prirodno okruženje gdje fizikalni, biološki i kemijski procesi koji se istovremeno odvijaju određuju kakvoću vode. Izradom različitih numeričkih modela omogućuje se pouzdanije praćenje kakvoće vode. Za izradu modela predviđanja kakvoće vode primjenom unaprijednih neuronskih mreža odabrana je postaja Donji Miholjac na Dravi za razdoblje od 1993. do 2016. godine. Pri odabiru parametara koji će se modelirati i njima pripadajućih ulaznih parametara korištena je korelacijska analiza (Pearsonova korelacija, korelacija udaljenosti). Klaster analizom i podjelom cijelog skupa na male, srednje i velike vode pomoću percentila definirani su ekstremni hidrološki scenariji. Ovaj će rad pokazati mogućnost predviđanja otopljenog kisika pri različitim hidrološkim scenarijima primjenom modela unaprijednih neuronskih mreža. Na slivu se pretpostavlja primjena određenih mjera poboljšanja. Cilj ovog modela je omogućiti bolje praćenje stanja kakvoće vode. Time bi se dala ocjena učinkovitosti pretpostavljenih mjera poboljšanja pri različitim hidrološkim scenarijima., Rivers are a complex natural environment with the physical, biological and chemical processes occurring simultaneously, determine the water quality. Various numerical models provide reliable water quality monitoring. Feedforward neural network models were applied for creating water quality prediction models of the river Drava at Donji Miholjac during the 1996 - 2016 period. Correlation analysis (Pearson Correlation, Distance Correlation) was used for selecting the parameters that will be modeled and their corresponding input parameters. Extreme hydrological events are defined by cluster analysis and distribution of whole dataset into small, medium and large water flows. This paper will show the possibility of predicting dissolved oxygen in different hydrological scenarios using the feedforward neural network model. The implementation of certain improvement measures is assumed on the basin. The model aims to provide a better monitoring of water quality. This would provide assessment of the effectiveness of the assumed improvement measures at different hydrological scenarios.

The territory of Croatia is located in a highly prone earthquake area with the threat from earthquakes producing ground accelerations ranging from 0.10g to 0.38g. More than half of the Croatian territory (56.22%) with more than one third (1, 633, 529) of the total Croatian population is characterized as a zone with a high risk of occurrence of earthquakes. In order to reduce primary catastrophic consequences of earthquakes, certain preparedness and emergency procedures have to be organized in the event of and prior to an earthquake. Earthquake risk refers to the expected losses to a given element at risk, over a specified future time period. Risk may be measured in terms of expected economic loss, or in terms of number of lives lost or the extent of physical damage to property. Kindergartens in Osijek have an important role in the educational process. For each of the 21 kindergarten buildings, the properties related to the year of construction, height, type of structure, total area, etc., are given. Since they mostly have only a base floor and are built as reinforced concrete buildings, they may serve as emergency shelters after earthquake events. Therefore, a complete strategy for evaluating their capability to face probable earthquakes has to be provided. The aim of the article is to determine seismic vulnerability of kindergarten buildings - the degree of loss to a given element at risk resulting from a given level of hazard, defined as a ratio of the expected loss to the maximum possible loss on a scale from 0 to 1, where 0 means without damage and 1 means collapse of building.

Measurements made in the past few decades undeniably indicate climate change. The most visible sign of global climate change is air temperature, while less visible indicators include changes in river water temperatures. Changes in river temperature can significantly affect the environment, primarily the biosphere. The physical, biological and chemical characteristics of the river are directly affected by water temperature. The estimation of this relationship presents a complex problem. Although river temperature is influenced by hydrological and meteorological factors, the purpose of this study is to model daily water temperature using only one known parameter – mean air temperature. The relationship between the daily mean air and daily water temperature of the River Drava is analysed using linear regression, stochastic modelling or nonlinear regression and multilayer perceptron (MLP) feed-forward neural networks. The results indicate that the MLP models are much better models and they can be used for the estimation and prediction of daily mean river temperature.

The movement of the flood wave along the river bed is a complex process, not only because of this that the flow varies in time, but because the water course hydraulic and geometric characteristics of the river bed is very variable. Propagation of water waves is the process that defines the characteristics of flood waves on a down stream section of the river flow on the basis of known flood waves in the upstream section. The main problem with the forecasts of flood wave output is in the mathematical description of the transformation input flood wave. The wave propagation can be illustrated by the space and time functions of discharge and depth. Flood routing is the determination of flow conditions in a river, given its initial state, the stream morphology and an inflow hydrograph. Flood- routing procedures may be classified as either hydrological or hydraulic. Hydrological methods use the principle of continuity and a relationship between discharge and the temporary storage of excess volumes of water during the flood period.Hydraulic methods of routing involve the numerical solutions of either the convective diffusion equations or the one–dimensional Saint–Venant equations of gradually varied unsteady flow in open channels. An example of a simple hydrological flood-routing technique used in natural channels is the Muskingum flood- routing method. In this paper for predicting output flood wave on the section of the Drava River (in the lower reaches) two methods are applied: Muskingum methods and a modified Kalinin-Milyukov method. The Muskingum method belongs to the class of hydrologic routing methods. Hydrologic routing models involve only storage and flow rates of floodwater and aim to capture the wave dynamics through their parameters. The Kalinin and Milyukov method is especially appealing for flow routing, because both model parameters, the number of linear reservoirs in the series and the time constant (or storage coefficient of the linear storage-outflow relationship of each reservoir in the cascade) are related to the channel’s characteristics and hydraulic conditions.

Flood protection in the Republic of Croatia is a very complex process due to the vast hilly and mountainous areas with high rainfall intensities, large lowland river valleys, cities and valuable goods of the potentially affected areas, and lack of built and maintained protective systems. It is estimated that flooding potentially threatens about 15% of state land territory of which the greater part is now protected with different levels of security. Natural floods which occur in Croatia can be divided into five basic groups: river flooding due to heavy rains and/ or rapid melting snow and torrential flooding of small streams due to the short-term high-intensity rainfall, flooding in karst areas due to heavy rains and/or rapid snow melt and insufficient capacity of permeable natural sinks, flooding inland waters in lowland areas, glacial floods. Flood protection measures to be implemented are: structural measures and non-structural measures.