Shuichi Ikebuchi

Evaporation from Lake Biwa

Abstract An experiment has been under way since 1985 to determine the best method to estimate the evaporation from Lake Biwa. A comparison of direct measurements by the eddy correlation method and large evaporation pan method, and the indirect estimation by aerodynamic method and heat balance method was made. As a result, the estimation of evaporation from Lake Biwa was carried out for a long continuous time by the bulk transfer method, based on measurements of surface water temperature, wind velocity, air temperature and humidity at about 6 m above the lake surface, made on a platform 200 m offshore. The annual variation of evaporation was found to be greater in autumn and winter seasons, from September to March, than in spring and summer seasons from April to August.

Short‐term rainfall prediction method using a volume scanning radar and grid point value data from numerical weather prediction

A physically based short-term rainfall prediction method, which uses a volume scanning radar, is extended so that it utilizes grid point values from a numerical weather prediction model as supplementary information. The original short-term prediction method mainly consists of a conceptual rainfall model that can predict rainfall distribution, particularly over mountainous regions, in a qualitative sense. On the other hand, the grid point values from the numerical weather prediction model, the Japan Spectral Model developed by the Japan Meteorological Agency, are operationally distributed as the grid point value (GPV) data. In the original short-term prediction method the three-dimensional wind field as well as initial distributions of the air temperature and water vapor were identified using topography and upper air observations. In the extended method, however, in identifying those initial values, the information from the GPV data is used instead of the upper air observations in order to accommodate large differences in temporal and spatial resolution between radar information and upper air observations. It is noted that this extended method does not use predicted GPV rainfall data. The conceptual rainfall model plays the role of bridging the gap between radar information and numerical weather prediction scales. This extended method is applied to a rainfall event which occurred in the bai-u season (one of the rainy seasons of Japan) in July 1994. Results show that for the extended lead time of three and four hours, prediction of the expanding rainfall area was improved.

A stochastic approach to short-term rainfall prediction using a physically based conceptual rainfall model

An improved method for short-term rainfall prediction is presented. A previously proposed deterministic rainfall prediction method for real-time hydrologic applications is extended to a stochastic method. This method mainly consists of a physically based conceptual rainfall model that includes water balance and thermodynamics. The important element in this method is the translation of radar data to the model parameter of the conceptual model, which is incorporated into the numerical scheme of the mesoscale model. The extended Kalman filter is used as a state estimator to update the model parameter of the conceptual model with new radar data and with forecasts from a numerical weather prediction model. The performance of the stochastic method is examined for a radar observation area that includes a mountainous region with a rainfall event that occurred along a front. The stochastic method performed better than the deterministic method.

Advanced use into rainfall prediction of three-dimensionally scanning radar

A computational method for the determination of rainfall distribution for applications in short term rainfall prediction is presented here. The method is strongly influenced by the experience gained from the observation and analysis of data gathered on a heavy rainfall event in 1986 that occurred during the Baiu Season in Japan. The method is based on the concept that rainfall occurs as an interaction between an instability field, appropriately modeled, and a field of water vapor under the influence of topography. The results from this computational method showed good agreement with the temporal variation in the rainband that moved across the observation field in 1986. Towards determination of the parameters in the computational model, another method for the determination of the rainfield is also developed. This second method determines the rainfall distribution from estimation of the conversion rate of water vapor to liquid water through use of data from a three dimensional scanning radar. The results are consistent with those obtained from the first method.

Basinwide flood control system by combining prediction and reservoir operation

Japan has traditionally performed flood prevention through the construction and use of dikes, storage reservoirs, and basins which are costly and time consuming options. Another non-structural option is to operate the flood control system appropriately with a view to reducing flood damage. In this paper, a flood control system combining the runoff prediction model in the whole river basin with the reservoir operation is discussed. Different models of the runoff process are introduced in order to compare their accuracies and the computational time for the flood forecasting system. The reservoir operational rule is formulated in terms of fuzzy inference theory. Historical data are applied in a case study for verification of the proposed theories.

TREE-RING WIDTH AND STABLE CARBON ISOTOPE COMPOSITION OF JAPANESE CYPRESS IN THE LAKE BIWA AREA, CENTRAL JAPAN, AND THEIR HYDROLOGIC AND CLIMATIC IMPLICATIONS

Chronologies of tree-ring width and stable carbon isotope composition of Japanese cypress were developed to help reconstruct a 300-year record of past hydrologic and climatic environments in the Lake Biwa area, central Japan. Site chronologies were built with 37 trees for ring width and four trees for carbon isotope composition, respectively. Correlation analysis with monthly climatic data revealed that radial growth of the trees is related to temperature in early spring, precipitation (or number of precipitation days) in early summer and precipitation in previous-year summer to autumn. Tree-ring cellulose carbon isotopic composition is correlated most significantly with the number of precipitation days in early summer months. Consequently, a chronology of the number of precipitation days in May was reconstructed by multiple regression analysis with ring-width and carbon-isotope predictors and was validated by comparison with the recent observed record.

Severe rainfall prediction method using artificial intelligence

A severe rainfall prediction method using artificial intelligence, which can forecast the time series variation and spatial distribution of severe rainfall is developed for supporting flood control management. This method is unprecedented in the sense that severe rainfall is predicted mainly based on the physical processes of cloud development, qualitative reasoning, and the model based reasoning.

Knowledge-Based System for Reservoir Operation during Low Flows Utilizing Weather Forecast Information

A model for long range and real time reservoir operations is developed, considering the medium and long range weather forecast provided by the meteorological agency. The reasoning employed by the reservoir operator to make the appropriate decision on the reservoir operations, in the presence of uncertainty and inevitable errors in the forecast, is modeled through a rule-based scheme. A fuzzy inference procedure is used to evaluate the rules and produce the control output.

Severe Rainfall Prediction Method Using Artificial Intelligence Based on the Knowledge of Convective Cloud Processes

The severe rainfall prediction method using artificial intelligence which simulates the act of weather forecast experts who are well versed in the local weather is proposed in order to support the flood control.Under the condition that the system does not take the place of the numerical weather forecast, the objectives of development of the system are as follows; i) to forecast severe rainfall in fine grid scale by consideration of subgrid phenomena which are difficult to be expressed in numerical models, and ii) to make a real-time explanation of the important causes of severe rainfall to river managers.Then, the subsystems of the SRAI are developed and discussed in terms of their performance. The accuracy of the system was 80% for forecast of two hours ahead rainfall over 10mm/hr. The system shows the important initial conditions for severe rainfall effectively using backward reasoning.

Knowledge Acquisition and Qualitative Reasoning for Flood Control

In this paper, we will develop an inference system for supporting flood control by using qualitative reasoning which is one of the methods of artificial intelligence. First we will refer to difficulties which occur when we apply the qualitative reasoning to flood control as well as general difficulties inherent to qualitative reasoning. We will propose how to solve the former type of difficulties. In our system, a method of controlling inference flow eliminates these difficulties. Finally, we apply the system to a real flood caused in the Managawa river.