JP3514564B2 - Mirage occurrence prediction method and prediction device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mirage occurrence prediction method, and more particularly to a method and apparatus for accurately predicting the occurrence of a mirage from tomorrow to at least 7 days ahead by using a statistical method.

[0002]

2. Description of the Related Art According to the literature, mirage is a phenomenon that has been observed since ancient times along the coast of the Sea of Okhotsk in Hokkaido, along the coast of the Sea of Japan from Tohoku to the San'in region, or around the Pacific coast including Chiba, Wakayama and Kagoshima prefectures. , The cause of its occurrence has been investigated and discussed for a long time, but it has not been predicted in advance about the occurrence of mirage. The mirage in Uozu City, Toyama Prefecture, which appears every year from the beginning of March to the end of June, is famous worldwide, and the mirage research group in Uozu has observed the mirage phenomenon here for the past 20 years. Valuable records that describe the meteorological conditions, etc. are kept carefully. However, the statistical relationship between the occurrence of mirages and meteorological conditions has not yet been clarified, and it is the current situation that the occurrence of mirages must be relied on past experience and intuition.

[0003]

However, as described above, it is impossible to expect a highly accurate and appropriate prediction with the method that depends on the experience and intuition of a person to predict the appearance of a mirage. Not only is it lacking in reliability even if it is used as a tourism resource for rural development and town development,
It is inevitable that a person who predicts the date and time of occurrence will be a heavy burden.

In view of such circumstances, an object of the present invention is to clarify the statistical relationship between the occurrence of a mirage and meteorological conditions, and to accurately and easily predict the presence or absence of a mirage.

[0005]

Means for Solving the Problems As a result of intensive research to achieve the above-mentioned object, the present inventor has applied a statistical method to elucidate mirage observation record data for many years, and
The inventors have found that the intended purpose can be achieved by selecting and using specific meteorological data for the analysis item, and have completed the present invention.

That is, the present invention analyzes the observed meteorological data relating to the occurrence of a mirage in the past, derives a prediction formula for determining the presence or absence of a mirage, and uses this to analyze the predicted meteorological data in the target area. By comparing the analysis value with the analysis value of the past observed meteorological data, it is possible to determine whether or not a mirage has occurred at least 7 days ahead. Further, in carrying out the above method, By inputting the forecasted meteorological data after the present in the target area into the arithmetic processing system that has been learning-controlled so that the observed meteorological data related to the past mirage occurrences can be input and analyzed / determined whether or not the mirage occurred The gist is a method of predicting the occurrence of a mirage, which is characterized by outputting the prediction information of the occurrence of a mirage.

Furthermore, a computing device for analyzing the past observed meteorological data relating to the occurrence of the mirage and deriving a prediction formula for determining the presence or absence of the mirage, and the computing device for observing the past observed meteorological data and the target area The gist of a mirage generation prediction device is an input device that inputs predicted weather data, and an output device that outputs a determination result of the presence or absence of a mirage obtained by analyzing the input data by a prediction formula.

[0008]

BEST MODE FOR CARRYING OUT THE INVENTION In carrying out the present invention, it is first necessary to select a meteorological item closely related to the presence or absence of a mirage. Therefore, a close examination of the mirage occurrence records and meteorological data found in the past literature shows that the common atmospheric conditions in each target area are atmospheric layers with high temperature change rates and different densities in the air layer contacting the sea surface or the ground. Is formed, the atmosphere is stable and this state is difficult to collapse,
It was confirmed that the mirage is likely to occur and is easy to observe when the weather is fine and the visibility is good.

As specific meteorological items related to the above-mentioned meteorological conditions, generally, wide area meteorological data such as a mobile high pressure center position, a thermal ridge existing area and a thermal trough existing area, and a difference between sea surface or surface temperature and air temperature. This includes narrow-range meteorological data such as the temperature decrease rate in the lower atmosphere, the stability index of Showerter, maximum temperature, weather, sunshine hours, rainfall, wind speed and direction, and winter snowfall.

In practicing the present invention, it is possible to use all of the above-mentioned meteorological items, but the items that are less effective or have a high internal correlation are searched for and reduced, and as few meteorological item data as possible are used. Is preferred. The weight of each of these meteorological items is not always constant and varies depending on geographical conditions such as the topography of the target area. Therefore, if this work is appropriate, not only the time required for calculation but also the accuracy of calculation should be improved. Is also possible.

For example, for mirages (hereinafter abbreviated as spring-summer mirages) that are observed in the season when the sea surface temperature or the surface temperature is lower than the air temperature, there are two locations, a mobile high pressure center position and a thermal ridge existence area. Selected from wide area meteorological data, difference between sea surface or surface temperature and temperature, maximum temperature, Shoalter's stability index, weather, sunshine hours, wind direction at maximum wind speed and winter snowfall 2
It is preferable to use a total of 4 to 9 meteorological item data consisting of 7 to 7 narrow area meteorological data.

Regarding mirages (hereinafter abbreviated as winter mirages) observed in the season when the sea surface temperature or the surface temperature is higher than the air temperature, the movable high pressure central position, the thermal trough existing area, and the thermal ridge existing area. 6-8 in total from regional wide area meteorological data and difference in sea surface or surface temperature and temperature, stability index of shoalter, weather, sunshine hours, and wind direction at maximum wind speed. It is preferable to use the weather item data of.

The reasons for selecting and using the above-mentioned meteorological items are as follows. 1. Rate of temperature change in the air layer in contact with the sea surface or surface: If the temperature of the sea surface or the surface of the earth is low and the temperature is high, or vice versa, the temperature change rate in the vertical direction of the air layer becomes large and a mirage tends to appear. That is, the above-mentioned spring-summer mirage occurs in the season when the daytime temperature rises and becomes higher than the seawater temperature, and
The probability of occurrence is higher on days when the maximum temperature is higher and the temperature change rate of the air layer in contact with the sea surface is higher. The presence of a thermal ridge (a temperature peak, where the isotherm rises to the north in the weather map) near the target area is extremely convenient for the appearance of such weather conditions. On the other hand, when the winter mirage occurs, the temperature drops and becomes considerably lower than the seawater temperature. The presence of a thermal trough near the target area (a valley of temperature that means that the isotherm hangs to the south in the weather map) is a good indicator of the temperature drop and is preferable for the emergence of a winter mirage. Depending on the target area, instead of the sea surface temperature, it is also possible to use snowfall that melts snow and affects the sea surface temperature and the amount of rainfall that melts the snow.

2. Atmospheric stability: Atmospheric conditions are generally stable in high pressure areas. The Schwarter Stability Index (SSI) is usually used as a measure of the stability of the upper atmosphere. The SSI value is 850h
It shows the temperature difference from the surroundings when the air mass on the Pa side is adiabatically lifted to an altitude of 500 hPa. When this value is large (that is, the lifted air is colder), it is in the high pressure area and the upper layer. It can be said that the atmosphere is stable.
In addition, as a measure for the stability of the lower atmosphere, the lower temperature decrease rate obtained by the following equation is used. Lower air temperature decrease rate (℃ / 100m) = (T ₁₀₀₀ −T _s ) / Z ₁₀₀₀ However, T ₁₀₀₀ : 1,000 hPa surface temperature T _s : Ground temperature Z ₁₀₀₀ : 1,000 hPa surface altitude Normally, in the atmosphere Then, the temperature becomes lower toward the sky, and the rate of decrease is about 0.6 to 1.0 degrees per 100 meters. If the rate of decrease is higher than this, the atmosphere is unstable. Conversely, when the temperature is higher in the sky, it is called "reversal", and in this case, the atmosphere is very stable.

3. Weather: First of all, it is preferable that the weather is clearer than cloudy for the generation and observation of the mirage. In addition, in some regions, the weather of the day before and the day before the day is related to the occurrence of the mirage. The wind speed is not so strong and not so weak. The wind direction differs depending on the region, but there is a favorable wind direction in correlation with the topography. In regions facing the sea or lakes, the temperature of the sea or water is low, and on land, there is a high probability that a mirage will appear when the weather is sunny and the temperature is high. At such times, a local sea breeze or lake breeze will blow. From such a viewpoint, the wind direction and speed are also indicators for predicting the occurrence of a mirage.

In carrying out the present invention, both the matters to be clarified (whether the mirage appears or not) and many matters related thereto (the above-mentioned meteorological item data) are converted into qualitative data and quantified. It is convenient to use a multivariate analysis method called class II. In carrying out the analysis, first, regarding the past observed meteorological data, the presence or absence of a mirage is used as the objective variable, and the specific meteorological item data is used as the explanatory variable, and the relationship between them is investigated, and the objective variable is strongly influenced. Discover the explanatory variables and derive the prediction formula for determining the presence or absence of a mirage. Next, using this prediction formula, the forecasted weather data from tomorrow onward will be analyzed, and the obtained results will be collated with the analyzed values of the past observed weather data to determine whether or not a mirage has occurred.

The past observed meteorological data used in the above is mainly obtained from various weather-related materials under the jurisdiction of the Meteorological Agency, such as the Japan climate table, printed weather maps, high-level meteorological observation data, and local monthly weather report. The data is created on the basis of narrow-range or wide-area weather data such as the Japan Meteorological Agency's GPV, weekly forecasts, and forecast weather maps. In this case, in order to improve the discrimination accuracy, it is desirable to process the GPV into an expected value of 5 km mesh and use it as meteorological data in a narrow area.

As described above, in the present invention, since the occurrence of the mirage is predicted by using the statistical method, it is possible to make the prediction with higher accuracy than the method which has conventionally depended on the intuition and experience of the person. Moreover, it is now possible to provide forecasts for at least 7 days ahead, as well as tomorrow's forecasts.
As a result, the mirage will soon be fully utilized as a tourism service resource.

[0019]

EXAMPLES Next, the present invention will be described in more detail with reference to examples. Example 1 Table 1 is a variable table showing objective variables and eight explanatory variables that are applied in predicting the presence or absence of a mirage tomorrow. Using this as a quantification type II analysis model, 282 samples of observation records of existing spring-summer mirages in Toyama Bay (observation data from April 1976 to June 1994) (of which 10 are independent for verification) (Used as data) was analyzed.

[0020]

[Table 1]

The above eight explanatory variables, namely, normal temperature deviation of maximum temperature, sunshine hours, SSI, numerical data such as snowfall depth, character data such as weather and wind direction, mobile high pressure center position, presence of thermal ridge All graphic data such as areas were converted into category data. Fig. 1 shows the categorization of the mobile high pressure center position, and Fig. 2 shows 850 hPa.
It shows the categorization of the surface thermal ridge existing area. The categorical classification of quantitative data and character data is summarized in Tables 2 and 3.

As described above, by inputting the variables converted into categorical data into the computer arithmetic processing system, the partial correlation coefficient and range representing the strength of influence of each explanatory variable and the category Scores and correlation ratios (coefficients representing the strength of influence on all explanatory variables) were calculated. The partial correlation coefficients are shown in Table 1, and the category scores are shown in Tables 2 and 3, respectively.

[0023]

[Table 2]

[0024]

[Table 3]

Needless to say, the quantification type II is a method of giving a score to each category of the explanatory variables according to the degree of contribution to the discrimination of the group of objective variables, and this score is the above category score. . The value obtained by adding the category scores of the applicable categories for each sample is called the sample score, and from this value it is determined which group each sample belongs to in the objective variable (here, mirage occurrence and mirage occurrence). Determine.

FIG. 3 is a graph showing the frequency distribution of sample scores calculated for the above 270 samples. In the figure, the mountain on the right side represents the first group of objective variables (with mirage occurrence), and the mountain on the left side is the second group (without mirage occurrence).
Is represented. The fact that the discrimination graph is clearly divided into two peaks in this way supports the good calculation accuracy.

As is clear from Tables 2 and 3, and FIG. 3, good results were obtained in terms of calculation accuracy, with a correlation ratio of 0.7952 and a discriminant predictive ratio of 91.5%.

Then, as described above, when 12 pieces of independent data for verification were used for verification, one of the 12 samples had a miss, and the discrimination accuracy was 0.92 (= 11/12). This value is almost the same as the above-mentioned discriminant predictive value.

Example 2 A method for predicting the occurrence of a mirage up to 7 days ahead was examined. As the predictive data, the data available in the routine were used, and the four explanatory variables shown in Table 4 were used. The conversion of each explanatory variable into categorical data is the same as in the first embodiment.

[0030]

[Table 4]

Here again, the quantification type II is used for the analysis method,
Observation record of the same spring-summer mirage as in Example 1 (April 1976)
282 samples of observation data from the month to June 1994) were analyzed. The correlation ratio obtained for this analytical model was 0.7678, and the discriminant predictive value was 88.2%. It can be said that good accuracy was obtained despite the small number of variables.

The procedure for using the above calculation result for forecasting is as follows. 1) Enter the daily forecast values up to 7 days ahead on the forecast worksheet shown in FIG. The weather and maximum temperature are based on the Toyama Weekly Forecast of the Local Meteorological Observatory and use their own modified values. The central location of the mobile high pressure and the area where the thermal ridge exists are determined by looking at the weekly forecast support map. 2) Determine the category for the forecast value from Table 5, read the category score of that category, and enter it on the worksheet. 3) Sum up the category scores of four variables on the worksheet to obtain the sample score. 4) Determine the presence or absence of mirage (with reference to the discriminant midpoint) based on the sample score.

[0033]

[Table 5]

[0034]

[Table 6]

It is needless to say that it is possible to automatically output the presence or absence of a mirage by using an arithmetic device, not limited to the method using the worksheet.

Next, the above method was applied and verified for two months (total 61 samples) in April and May 1995. The results are shown in Table 6, and the deviation was 8 out of 61 samples, and the discrimination accuracy was 0.87 (= (6 + 4
It was 7) / 61).

As for the method of predicting the occurrence of the winter mirage, the same method as described above can be used for the determination.

[0038]

As described above, the present invention is a method and apparatus for statistically analyzing past meteorological data and discriminating the occurrence of a mirage based on the statistical data. Therefore, reliable prediction can be made quickly. Can do at least 7
It has a remarkable effect that the information about the occurrence of mirage up to the future can be accessed anytime and anywhere through the communication service network.

[Brief description of drawings]

FIG. 1 is a diagram showing category division of a mobile high pressure center position.

FIG. 2 is a diagram showing category division of a thermal ridge existing area of 850 hPa surface.

FIG. 3 is an example of a graph showing a frequency distribution of sample scores.

FIG. 4 is an example of a weekly forecast worksheet.

─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-9-49884 (JP, A) JP-A-7-225284 (JP, A) JP-A-6-300860 (JP, A) JP-A-2- 221997 (JP, A) JP 7-260946 (JP, A) JP 5-134055 (JP, A) JP 4-253282 (JP, A) Fukiyama Naoki, Mochizuki Higashi, Asami Taizo, "Weather Of Lightning Prediction Using Factors ", The Japan Society for Electrical Discharge and High Voltage Joint Research Workshop, Japan, The Institute of Electrical Engineers of Japan, October 26, 1995, ED-95-197-207, p. 23-32 (58) Fields surveyed (Int.Cl. ⁷ , DB name) G01W 1/00-1/18 G06F 17/60-19/00 JISST file (JOIS)

Claims (9)

(57) [Claims] 1. A forecast formula for determining the presence or absence of a mirage is derived by analyzing past observed meteorological data related to the occurrence of a mirage, and the forecasted meteorological data in the target area is analyzed using this, and the analysis value is obtained. A mirage generation prediction method characterized by determining whether or not a mirage has occurred at least 7 days ahead by comparing with the analysis value of past observed meteorological data. 2. The forecasted meteorological data after the present in the target area is input to an arithmetic processing system that is learning-controlled so that past observed meteorological data related to the occurrence of a mirage can be input and analyzed / determined whether or not the mirage occurs. According to the above, the mirage occurrence prediction method is characterized by outputting the mirage occurrence prediction information for at least 7 days ahead. 3. The meteorological data according to claim 1 or 2,
Rate of temperature change in the air layer in contact with the sea surface or surface of the target area,
A method for predicting the occurrence of mirages, which consists of a combination of meteorological item data related to atmospheric stability and weather. 4. The meteorological item data according to claim 3 includes a mobile high pressure center position, a thermal ridge existing area and a thermal trough existing area, a difference between sea surface or surface temperature and air temperature, a temperature decrease rate in the lower atmosphere, and a shoaler's data. Stability index, weather, daylight hours, rainfall, wind speed and wind direction mirage occurrence prediction method. 5. The meteorological item data according to claim 3, wherein the difference between the sea surface or surface temperature and the temperature, the normal deviation of the maximum temperature,
At least two meteorological item data selected from low-temperature deceleration of the lower atmosphere, stability index of shoalter, weather, sunshine duration, wind direction at maximum wind speed and narrow range meteorological data of winter snowfall, and mobility A mirage generation prediction method comprising a combination of at least two meteorological item data selected from wide area meteorological data of a high pressure center position, a thermal ridge existing area, and a thermal trough existing area. 6. An arithmetic device for deriving a prediction formula for analyzing whether or not a mirage has occurred by analyzing past observed meteorological data relating to the occurrence of a mirage, and a past observation meteorological data and a prediction in a target area in this arithmetic device. A mirage generation prediction device comprising an input device for inputting meteorological data and an output device for outputting a determination result of presence or absence of a mirage obtained by analyzing the input data by a prediction formula. 7. A mirage outbreak prediction in which the meteorological data according to claim 6 is a combination of a temperature change rate of an air layer contacting the sea surface or the surface of the target area, atmospheric stability, and meteorological item data related to weather. apparatus. 8. The meteorological item data set forth in claim 6 includes a mobile high pressure center position, a thermal ridge existing area and a thermal trough existing area, a difference between sea surface or surface temperature and air temperature, a temperature decrease rate of the lower atmosphere, and a shoaler value. Stability index, weather, sunshine hours, rainfall, wind speed and wind direction mirage occurrence prediction device. 9. The meteorological item data according to claim 6, wherein the difference between the sea surface temperature or the surface temperature and the temperature, the average deviation of the maximum temperature,
At least two meteorological item data selected from low-temperature deceleration of the lower atmosphere, stability index of shoalter, weather, sunshine duration, wind direction at maximum wind speed and narrow range meteorological data of winter snowfall, and mobility A mirage generation prediction device comprising a combination of at least two meteorological item data selected from wide area meteorological data of a high pressure center position, a thermal ridge existing area, and a thermal trough existing area.