Evaluating stream water quality through land use analysis and Nitrogen budget approaches

K.P. Woli

Graduate School of Agriculture , Hokkaido University

There is growing concern on environmental degradation due to nitrogen (N) pollution, and intensification of agriculture and livestock husbandry practices has made a significant contribution to N pollution in stream water, particularly nitrate contamination. The objective of this investigation was to evaluate the quality of stream water by analysis of land use in drainage basins and by N budget approaches. The impact of land use proportions in drainage basins on N concentration in stream water was evaluated in seven catchment areas in Hokkaido for various land use systems. Water sampling was carried out from altogether 255 sites in seven areas during the base flow events, and nitrate-N (NO₃-N) concentration was measured. Results indicated that the proportion of uplands (including grasslands) in drainage basins had significant positive correlation, forests had significant negative correlation, and paddy fields had no correlation with NO₃-N concentration in stream water. The regression slopes of the relation between upland proportion and NO₃-N varied according to land use systems, being highest in intensive livestock farming (0.040), intermediate in mixed agriculture-livestock farming (0.020-0.030), and lowest in dairy cattle, horse farming, and mixed paddy-upland (0.0052-0.015) areas. The slopes thus acted as impact intensity of uplands in raising NO₃-N concentration; therefore, they were defined as the impact factor (IF) of water quality. To quantitatively clarify the variation in IFs of water quality for different areas, regression analysis was performed between IFs and independent parameters such as livestock and human densities, and N loading variables such as chemical and compost fertilizer N applied, human disposed N, livestock disposed N, and cropland surplus N, which were estimated by N budget approaches. Simple linear regression indicated that cropland surplus N (CSN) had the best correlation with the IFs (r=0.93, p<0.01), being the regression model as IF=1.09x10^-4＊CSN+0.017. By using this regression model, IFs for all cities, towns, and villages in Hokkaido were predicted using the reported information on CSN. The pattern of predicted IFs was very close to that of NO₃-N concentration for all major rivers in Hokkaido , as reported in a previous study. I estimated NO₃-N concentration for all those sampling sites in Hokkaido by multiplying the predicted impact factors by the proportion of uplands. The regression indicated that the predicted NO₃-N concentration significantly correlated with the measured NO₃-N concentration (r=0.62, p<0.001). In conclusion, higher proportions of upland in drainage basins enhanced NO₃-N concentration in stream water, and the impact intensity of land use on water quality was as large as the amount of CSN in respective drainage basins. A proper management of upland proportions in drainage basins and of N surpluses in cropland is very important in protecting stream water quality.

Experimental study on the time-dependent deformation and failure of mortar

K. N. Adhikari

Graduate School of Engineering , Hokkaido University

Detachments of concrete blocks from tunnel lining have often occurred recently in Japan . One of the mechanisms of tunnel lining failure is as follows (Fig. 1): (1) Yielding or creep deformation occurs in rock mass at sidewalls of the tunnel lining under almost constant concentrated vertical stress. The yielding might be creep failure without any trigger or induced by such triggers like pore pressure increase, etc. (2) The sidewalls expand and the lining is subjected to horizontal pressure, increasing slowly. (3) Compressive and tensile failure occurs in the cemented portion, the weakest part of the lining.

Considering that more cracks appear in the matrix (mortar) than the aggregates and that the invert of tunnels and lining of water tunnel are usually wet, the uniaxial compression, Brazilian, and CB tests for dry and saturated mortar specimens were carried out under several loading rates in order to clarify the effect of compressive/tensile loading rate on the time-dependent failure behavior of lining. According to several articles, the actual displacement rate of tunnel lining is roughly estimated to be 10^-9 to 10^-10 m/s. Considering the difficulty of application of lower loading rates, only 10^-8 m/s was used in this study. Time-dependent behavior of mortar in Brazilian and CB tests are first made in this research so far as to the author's knowledge. Creep tests were carried out for saturated specimens in pure water since the lining can be subjected to a constant load.

Multi-stage creep test on physical model simulating circular roadway was carried out for two models having different mixture ratios in air and water respectively to clarify the time-dependent deformation of surrounding rock mass.

The interaction between lining & rock mass has to be considered to investigate lining failure more precisely. Unfortunately, this could not be done in the limited time given to the author. However, it is hoped that such interaction will be clarified and more practical ways to predict lining failure will be found in near future.