The dynamics of water quality and phytoplankton population had examined by monthly sampling from the upper to the lower part of watergate in an artificial Shihwa Reservoir in which situated near newly cities and industrial complex on the west coast of Korea from January 1997 to December 1998. Among environmental factors, yearly average concentration of chl-a, TN and TP seemed to eutrophic or hypertrophic conditions that ranged 146.4∼245.8㎍/ℓ , 1.6∼2.7㎎ N/ℓ, 258∼448㎍ P/ℓ , 26.9∼80.7 ㎍/ℓ, 1.0∼2.4 ㎎ N/ℓ and 74∼239㎍ P/ℓ respectively. Water quality was extremely deteriorated to consistently accumulation into inner reservoir by load of pollutants from autochthonous and allochthonous until early July 1997 after embankment. Water pollution of Shihwa Reservoir was remarkble on the biological condition with largely persistent bloom of phytoplankton and increase rate of standing crops was 2.4/yr. The development trend of phytoplankton in water ecosystem were closely related to increse and decrease of physico-chemical factors and those scale seemed to control by nutrient contents. Inflow of seawater into reservoir to object of repair of water quality began on July 1997. This action was caused to variable of algal flora as well as improvement of water quality. As to see dominant species, composition of those composed to mostly freshwater algae before inflow of seawater such as Selenastrum capricornutum of green algae, Cyclotella atomus, C meneghiniana of diatom and Microcystis spp. of blue-green algae and the other hand brackish algae were dominated after inflow of seawater such as Chaetoceros dicipiens, Skeletonema costatum of diatom, Dinophysis acuminata, Gymnodinium mikimotoi, G. sanguineum, Gyrodinium spirale, Prorocentrum minimum of dinoflagellate and Eutreptiella gymnastics of euglenoid. Moreover, small flagellates including Chroomonas spp. of cryptomonad were abundant throughout the year. The cause of water deterioration during fill of the freshwater were complexly supported with extra and intra parameters. The variation pattern of phytoplankton were related to water temperature and salinity by inflow of seawater based to plentiful nutrients. The dynamics of phytoplankton were assessed to ecosystem that clearly condition of dominant by unique or a few algal species seasonally.

본 논문의 목적은 다목적 저수지의 홍수유입량 예측을 위한 방법으로 병렬다중결선의 계층구조를 가진 신경망이론에 의하여 홍수시 불확실한 비선형시스템의 특성을 같는 저수지 유입량 예측모형을 개발하는 것이다. 신경망이론을 이용한 예측모형의 개발을 위하여 역전파 학습알고리즘을 사용하였으며 역전파 학습알고리즘 사용시 흔히 대두되는 지역최소값 문제와 수렴속도의 향상을 위해서 최적화기법인 경사하강법을 이용한 모멘트법과 경사하강법과 Gauss-Newton 방법을 이용한