Estimation of nitrate transport in an unsatured zone on soils with different physical properties

Onečišćenje nitratima predstavlja važan ekološki problem, posebno u poljoprivrednim prostorima. Nitrati se ne mogu vezati na adsorpcijski kompleks tla, te zbog toga dolazi do njihovog ispiranja u dublju zonu tla ili podzemnu vodu. U svrhu ovog istraživanja uzeti su uzorci tla na pet različitih lokacija (grad Zagreb i okolica Zadra) te su određene karakteristike tla. Na neporušenim uzorcima određene su retencijske krivulje tla te vrijednosti hidrauličkih parametara tla (HYPROP). Za validaciju modela pronosa nitrata u kontroliranim uvjetima pratilo se ispiranje nitrata na kolonama s tlom na jednoj lokaciji (tip tla: Rendzina, lokacija: Baštica). U perkolatu su određeni nitrati. Na kolone je primijenjena otopina amonij-nitrata ekvivalentna količini gnojidbe od 100 kg N/ha, te se primjenom statističkih parametara (R2, RMSE) validirao adekvatan model. Nakon uspješne validacije modela na testnom području provelo se četiri različitih simulacija na temelju podataka o tlu na području grada Zagreba. U modelu su korišteni identični rubni i inicijalni uvjeti. Modeliranje se provelo tijekom jedne vegetacijske sezone pomoću Richardsove jednadžbe za tok vode i jednadžbe advekcijsko disperzivnog pronosa za nitrate uz identične rubne i inicijalne uvjete. Pogreška modela retencijskih krivulja (RMSE_TH) na svim profilima bila je < 0,01 cm3, te možemo zaključiti da je primjenjivani van Genuchten-Mualemov model bio uspješan, te su procjene hidrauličkih parametara pouzdane. Pronos nitrata uvelike je ovisio o hidrauličkim parametrima tla u našem slučaju, a općenito pronos nitrata ovisi o primjeni mineralnih i organskih oblika dušika te rasporedu i intenzitetu oborina.Nitrate pollution is one of the leading problems of today, especially in agricultural holdings. Nitrates cannot bind to the adsorption complex of the soil, which is why they leach into the deeper zone of the soil or groundwater. For the purpose of this research, soil samples were taken on five different locations (the city of Zagreb and the surroundings of Zadar) and soil characteristics were determined. Retention and hydraulic soil curves and values of hydraulic soil parameters (HYPROP) were determined on undisturbed samples. To validate the nitrate transport model under controlled conditions, nitrate leaching was monitored on soil columns at one location (soil type: Rendzina, location: Baštica). Nitrates are found in percolate. An ammonium nitrate solution equivalent to a fertilization amount of 100 kg N / ha was applied to the columns, and an adequate model was validated using statistical parameters (R2, RMSE). Five different simulations were made based on soil data with identical boundary and initial conditions. Modeling was performed during one growing season using the Richards equation for water flow and the advection-dispersive transport equation for nitrates under identical boundary and initial conditions. The error of the retention curve model (RMSE_TH) on all profiles was below 0.01 cm3, and we can conclude that the applied van Genuchten-Mualem model was successful, and the estimates of hydraulic parameters are reliable. Nitrate transport largely depended on the hydraulic parameters of the soil in our case, and in general nitrate transport depends on the application of mineral and organic forms of nitrogen and the distribution and intensity of precipitation