Meteorological data are often needed to evaluate the long-term effects of proposed hydrologic changes. The evaluation is frequently undertaken using deterministic mathematical models that require daily weather data as input including precipitation amount, maximum and minimum temperature, relative humidity, solar radiation and wind speed. Stochastic generation of the required weather data offers alternative to the use of observed weather records. The precipitation is modeled by a Markov Chain-exponential model. The other variables are generated by multivariate model with means and standard deviations of the variables conditioned on the wet or dry status of the day as determined by the precipitation model. Ultimately, the objective of this paper is to compare Richardson's model and the improved weather generation model in their ability to provide daily weather data for the crop model to study potential impacts of climate change on the irrigation needs and crop yield. However this paper does not refer to the improved weather generation model and the crop model. The new weather generation model improved will be introduced in the Journal of KWRA.

This paper presents the hydraulic analysis of the oxygen transfer through the air entrainment and the relationships between the efficiency of the oxygen transfer and the hydraulic parameters in the riparian riffles. Field survey on the pool-riffle formation of the river reach and the measurements of the oxygen transfer in the riffles were performed. Air entrainment occurred more frequently in the edged gravels rather than in the round and edgeless ones, and it was formed mainly from behind the trailing edges of the gravels. Oxygen transfer was found to be proportional to the flow velocity, the flow discharge, and the Froude number, but to be not closely related to the particle diameter. Average value of oxygen transfer in the riffles of study area was about 0.085, which shows good efficiency compared with results of smooth chute. Variation of the water level, which increases in proportion to the flow velocity and the flow discharge, seems to make the air entrainment more active, but has not been verified quantitatively. Relationships between the air entrainment and the variation of the water level must be considered in the further study.

Nonpoint source pollutants from agriculture are identified as one of the main causes of water quality degradation in the United States. The Hydrological Simulation Program-Fortran (HSPF) was used to simulate runoff, nitrogen, and sediment loads from an urbanizing watershed; the Polecat Creek watershed located in Virginia. Model parameters related to hydrology and water quality were calibrated and validated using observed hydrologic and water quality data collected at the watershed outlet and at several sub-watershed outlets. A comparison of measured and simulated monthly runoff at the outlet of the watershed resulted in a correlation coefficient of 0.94 for the calibration period and 0.74 for the validation period. The annual observed and simulated sediment loads for the calibration period were 220.9 kg/ha and 201.5 kg/ha, respectively. The differences for annual nitrate nitrogen ($NO_3$) loads between the observed and simulated values at the outlet of the watershed were 5.1% and 42.1% for the calibration and validation periods, respectively. The corresponding values for total Kjeldahl nitrogen (TKN) were 60.9% and 40.7%, respectively. Based on the simulation results, the calibrated HSPF input parameters were considered to adequately represent the Polecat Creek watershed.

Parameters and quantiles of the 2-parameter generalized Pareto distribution were estimated using the methods of regular moments, modified moments, probability weighted moments, linear moments, maximum likelihood, and entropy for Monte Carlo-generated samples. The performance of these seven estimators was statistically compared, with the objective of identifying the most robust estimator. It was found that in general the methods of probability-weighted moments and L-moments performed better than the methods of maximum likelihood estimation, moments and entropy, especially for smaller values of the coefficient of variation and probability of exceedance.