Browsing by Subject "Temporal variability"

Now showing 1 - 1 of 1

Numerische Modellierung und Simulation der räumlichen und zeitlichen Variabilität von Lachgasemissionen aus Agrarökosystemen
(2005) Huber, Stefan; Doluschitz, Reiner
The aim of this dissertation is to characterize the spatial and temporal variability of nitrous oxide emissions from agroecosystems by means of linking mathematical simulation models to Geographical Information-Systems (GIS). Specifically, this study tries to accomplish a methodological goal and a thematic goal. A general software framework for the linkage of agroecosystem-modells and GIS by employing object-oriented and component-based concepts is developed. As an example for the implementation of this framework, the agroecosystem-modell DNDC, the two-dimensional soil water model SWMS2d, and GIS are integrated to the new model Spatial DNDC. This new model is applied to the study of the spatial and temporal variability of nitrous oxide emissions from agroecosystems at different scales. The simulation of the emissions at 281 independent soil profiles from a 40 ha field in Michigan/USA for nine years shows a large temporal and spatial variability ranging from 0.35 to 4.21 kg N2O-N/ha/a. Except for three years the yearly emissions are always lognormal distributed. While comparing simulated and measured daily emissions cannot be regarded as satisfactorily, it can be shown that the median of the daily N2O-emission rates can be employed as a characteristic measure for the given site. The influence of lateral soil water movement on the emission of N2O is studied by employing SpatialDNDC on a dataset from Scheyern/Bavaria, which comprises measurement data for five sites along hill-slope transect for the year 1997. The simulated daily emission rates are very similar for the five sites and are in good agreement with the measurements. The temporal variability of the daily emission rates is largely shaped by the occurrence of nitrogen fertilization-events and following precipitation events. By looking at the yearly emissions a distinct, downslope-directed gradient can be seen with the highest emission of 6.87 kg N2O-N/ha/a at the highest hill position, and the lowest emission of 6.37 kg N2O-N/ha/a at the lowest hill position. This gradient can be explained by the soil water household which is largely influenced by a dry period in spring. Due to lateral water movement plants growing at lower positions have more water available for early-spring growth leading to higher water extraction during the dry period. Therefore the average soil water content, which is a major impact factor for N2O-emissions in SpatialDNDC, lower at the downslope positions as compared to the upslope positions. The modelling of N2O from agroecosystems in the North-China-Plain can merely be regarded as a test case for application of SpatialDNDC to larger regions and whole nations, respectively, since detailed input and validation data are missing. The three simulation studies show distinctively the two main problems of the regional usage of agroecosystem models: On the one hand detailed input data are missing leaving for the modeller only the option to make simplifying assumptions and thereby introducing great uncertainty into simulation results. On the other hand regional calibration and validation data are missing, which are crucial to the realistic depiction of variability within large study regions.