Method and system for estimating meteorological quantities6778929Abstract Estimation method for obtaining estimation results of meteorological quantities in a specified area during a specified future period, including steps of: provisionally creating a meteorological time-series model from historical data of the meteorological quantities observed in the specified area; adjusting parameters of the created time-series model on the basis of long-range weather forecast data for wider area, which contains future meteorological tendency relative to normal years, to adjust the created time-series model; and conducting simulation using the adjusted time-series model to obtain the estimation results. Claims What is claimed is: Description PRIORITY INFORMATION
TABLE 1
Standard
Parameter Average Deviation
.beta. 0.56 0.12
.mu. -0.01 0.36
.sigma. 1.83 0.26
A major cause of such fluctuation is occurrence of meteorological phenomena which are deviated from that of ordinary years such as "This summer is intensively hot" or "This winter is rather warm". On the contrary, when the parameters are fluctuated as much as the standard deviation of Table 1 (written as .DELTA..beta., .mu..DELTA., .DELTA..sigma.), how much do the E[T] and V[T] given by the [Expression 5] change? As a result of practical evaluation, the solution expressed by the [Expression 12] can be obtained: .vertline.E(.beta.+.DELTA..beta.,.mu.,.sigma.)-E(.beta.,.mu.,.sigma.).vertl ine.=0.25 .vertline.E(.beta.,.mu.+.DELTA..mu.,.sigma.)-E(.beta.,.mu.,.sigma.).vertlin e.=0.96 .vertline.E(.beta.,.mu.,.sigma.+.DELTA..sigma.)-E(.beta.,.mu.,.sigma.).vert line.=0 .vertline.V.sup.1/2 (.beta.+.DELTA..beta.,.mu.,.sigma.)-V.sup.1/2 (.beta.,.mu.,.sigma.).vertline.=0.28 .vertline.V.sup.1/2 (.beta.,.mu.+.DELTA..mu.,.sigma.)-V.sup.1/2 (.beta.,.mu.,.sigma.).vertline.=0 .vertline.V.sup.1/2 (.beta.,.mu.,.sigma.+.DELTA..sigma.)-V.sup.1/2 (.beta.,.mu.,.sigma.).vertline.=0.07 [Expression 12] However, E[T] and V[T] are values calculated for average temperature during the period up to Aug. 31, 2001 from Jun. 1, 2002. Moreover, each parameter can be estimated from the data during the period up to Dec. 31, 2000 from Jan. 1, 1971, while T.sub.0 is average temperature of a day of May 3, 2001. This result can be evaluated as follows. Here, it is assumed that average value of T (namely E) is changed by 1.degree. C. If it is also assumed that change of E is caused by fluctuation of .beta., fluctuation range of .beta. equal to about four times the .DELTA..beta. is required but probability of occurrence of the fluctuation as large as about four times the standard deviation is smaller than 1/10.sup.4. Meanwhile, when it is assumed that variation of E is caused by fluctuation of .mu., fluctuation range of .mu. is enough when it is equal to about .DELTA..mu. and probability of such fluctuation is about 30%. Accordingly, the fluctuation of .mu. assures the highest reliability for description of variation of E. As described above, the variation of E mostly results in the variation of .mu.. In the same way, the variation of V mostly results in the variation of .beta.. In other words, fluctuation of .sigma. can approximately ignored and a value estimated from the historical data of temperature in the past may be utilized. First, values of .beta., .mu., .sigma. are determined with the least squares method from the historical data of temperature in the past by paying attention to this physical property. Next, a value of .beta. is numerically solved again to satisfy the [Expression 11] based on the long-range weather forecast, and a value of .mu. can be solved again by substituting the value of .beta. after the adjustment to the [Expression 10]. Thereby, a set of parameters after the adjustment can be obtained. As described above, it is now possible to effectively acquire the physical information of the spatially wider area that probability of this summer to become cold summer is high by reflecting the long-range weather forecast on the parameters of the meteorological time-series model. However, the description developed here is never limited only to the Dischel model and a similar description can also be applied to the other proposed meteorological time-series models. Practical Aspects FIG. 2 is a flowchart illustrating an embodiment of the meteorological quantities estimation method of the present invention. As illustrated in FIG. 2, the estimation method of meteorological quantities of this embodiment comprises a data acquisition step 0201 for acquiring historical data of meteorological quantities 0211 as the historical data of meteorological quantities observed in the past, long-range weather forecast data announced from the Meteorological Agency 0212 and estimation condition 0213 as the condition for estimating meteorological quantities, a meteorological time-series model creation step 0202 for creating a time-series model of meteorological quantities by utilizing the information acquired in the data acquisition step 0201, a meteorological quantities simulation step 0203 for simulating meteorological quantities by utilizing the meteorological time-series model created in the meteorological time-series model creation step 0202 and a meteorological quantities estimation result output step 0204 for outputting the simulation result of the meteorological quantities simulation step 0203 as the meteorological quantities estimation result 0214. Moreover, the meteorological quantities time-series model creation step 0202 comprises a meteorological time-series model parameter estimation step 02021 for estimating parameter values of the meteorological time-series model from the historical data of meteorological quantities 0211 and a meteorological time-series model parameter adjustment step 02022 for adjusting the parameters estimated in the meteorological time-series model parameter estimation step based on long-range weather forecast data 0212. Next, an example of the historical data of meteorological quantities of FIG. 2 will be explained with reference to FIG. 3. The historical data of meteorological quantities 0211 of this embodiment includes date 031 and average temperature 032 of a day of each day at each observation point. In FIG. 3, for example, the average temperature of a day of Tokyo of Jan. 1, 1972 is 4.1[.degree. C.]. Next, an example of the long-range weather forecast data 0212 in FIG. 2 will be described with reference to FIG. 4. The long-range weather forecast data 0212 of this embodiment is composed of the forecast period 041, range of ordinary years 042 and probability distribution 043 of three grades including "below-normal", "near-normal" and "above-normal". FIG. 4 illustrates contents of announcement of long-range weather forecast announced from the Meteorological Agency. Next, an example of the estimation condition 0213 of FIG. 2 will be described with reference to FIG. 5. The estimation condition 0213 of this embodiment includes an estimation period 051 of average temperature of a day, an estimation point 052 and the number of times of simulation 053. FIG. 5 suggests that 10000 average time-series sequences of a day during the period up to August 31 from June 1 in Tokyo are generated with the simulation. Next, the data acquisition step 0201 of FIG. 2 will be described with reference to FIG. 3, FIG. 5 and FIG. 10. Since the estimation point 052 is selected to Tokyo in the estimation condition 0213 of FIG. 5, the historical data of meteorological quantities acquired from the historical data of meteorological quantities of FIG. 3 becomes the meteorological quantities in the past of Tokyo as illustrated in FIG. 10. However, in this embodiment, it is assumed that the forecast period 041 of FIG. 4 and the estimation period 051 of FIG. 5 are in the relationship that the forecast period 041 is matched with the estimation period 051 or the forecast period 041 includes the estimation period 051. Next, an example of the meteorological time-series model parameter estimation step 02021 of this embodiment will be described with reference to FIG. 6. This meteorological time-series model parameter estimation step 02021 is prepared to estimate the meteorological time-series model parameters from the meteorological data in the past. In the meteorological time-series model parameter estimation step 20021 of this embodiment, the parameters .beta., .mu., .sigma. of the Dischel model given by the [Expression 1] are estimated. As illustrated in FIG. 6, the meteorological time-series model parameter estimation step 02021 of FIG. 2 comprises an initialization step 0601, an observed temperature array creation step 0602, a normal year temperature array creation step 0603, an observation point array creation step 0604 and a regression analysis step 0605. In the initialization step 0601, the total years Y of historical data is stored by referring to the historical data of average temperature 0211 (FIG. 10). For example, since FIG. 10 includes the data of 31 years, the data Y=31 is stored. In the observed temperature array creation step 0602, the observed temperature array T[l] (l=1, . . . , 365*Y) is created by referring to the historical data 0211 (FIG. 10) of the average temperature under the condition that the day February 29 is skipped in the case of leap year. In the ordinary year temperature array creation step, the average temperature of the day in the same date during Y years, namely the ordinary year temperature array .THETA.[l](l=1, . . . , 365) is created based on the [Expression 13] from the observed temperature array: ##EQU9## Here, the regression formula such as [Expression 14] is prepared for the [Expression 1]: y=.beta.x+.mu. [Expression 14] However, .delta. is a residual in average 0. For this regression formula, observation points in the number of M=365*Y-1 are prepared based on the [Expression 15] introducing a regressand x and a regressor y: y.sub.n =T.sub.n+1 -.THETA..sub.n+1 x.sub.n =T.sub.n -.THETA..sub.n+1 (n=1, . . . ,M) [Expression 15] Here, the regression analysis is executed for these observation points to obtain .beta., .mu. and .sigma..sup.2 which is a variation of residual .delta.. Therefore, in the observation point array creation step, the regressor array Y[l](l=1, . . . , M) and the regressand array X[l](l=1, . . . , M) are created based on the [Expression 16] from the observed temperature array T[l] and ordinary year temperature array .THETA.[1]: Y[I]=T[I+1]-.THETA.[I+1] X[I]=T[I]-.THETA.[I+1] [Expression 16] In the regression analysis step, the values of .beta., .mu., .sigma..sup.2 are estimated based on the [Expression 17] derived with the least-squares method based on the X[l] and Y[l] created in the observation point array creation step. ##EQU10## Next, an example of the meteorological time-series model parameter adjustment step 02022 of this embodiment will be described with reference to FIG. 7. In this meteorological time-series model parameter adjustment step 02021, the adjustment process is executed to fetch the long-range weather forecast data to the parameters determined from the meteorological data in the past. As illustrated in FIG. 7, the meteorological time-series model parameter adjustment step 02022 of FIG. 2 is composed of a long-range weather forecast data storage step 0701, an estimation condition storage step 0702, a x1-x2 calculation step 0703, a .beta.-calculation step 0704 and a .mu.-calculation step 0705. In the long-range weather forecast data storage step 0701, the data probabilities for "below-normal" and "above-normal" are respectively stored to p.sub.1 and p.sub.2 and the upper bound and lower bound of the range of "near-normal" are respectively stored to T.sub.1 and T.sub.2 from the long-range weather forecast data obtained in the data acquisition step of FIG. 2. Moreover, the number of days of forecast period is calculated and is then stored to N, an ordinary year temperature array .THETA.[l](l=1, . . . , N) of the forecast period is stored from the ordinary year array .THETA.[l](l=1, . . . , 365), and the temperature of the preceding day of the forecast period is stored as T.sub.0 from the ordinary year temperature array .THETA.[l](l=1, . . . , 365). In the inverse function calculation step 0702 of the normal distribution, values of x.sub.1 and x.sub.2 are obtained from the [Expression 6]. In the .beta. calculation step 0703, the value of .beta. can be obtained from the [Expression 11]. However, the value obtained in the regression analysis step 0605 of FIG. 6 is used as .sigma.2. Moreover, since the value of .beta. can be obtained easily from the [Expression 11] using the Newton-Raphson's method, the detail method is not described here. In the .mu. calculation step 0704, the value of .mu. is obtained from the [Expression 18] derived from the [Expression 12] to obtain the value of .mu.. However, the value obtained in the .beta. calculation step 0702 is used for .beta.: ##EQU11## Next, an example of the meteorological quantities simulation step 0203 of this embodiment will be described with reference to FIG. 8. This meteorological quantities simulation step 0203 executes simulations based on the meteorological time-series model created in the meteorological time-series mode creation step 0202. As illustrated in FIG. 8, the meteorological quantities simulation step 0203 of FIG. 2 is composed of an estimation condition storage step 0801, a simulation initialization step 0802, a random number generation step 0803, a temperature estimation step 0804, an estimation period end determination step 0805, an estimation days addition step 0806, an estimation temperature time-series accumulation step 0807, a simulation end determination step 0808, and a simulation number addition step 0809. In the estimation condition storage step 0801, the number of days of estimation period is calculated and is then stored to N' and the number of times of simulation is stored to L to initialize the simulation number P to 1. In the simulation initialization step 0802, the estimation temperature time-series array S[l](l=1, . . . , N') is initialized to S[0]=T.sub.0, S[l]=0 (l=1, . . . , N') in order to initialize the estimation day counter C to 1. In the random number generation step 0803, a random number U is generated depending on the normal distribution of average .mu. variation .sigma..sup.2. However, the value acquired in the .mu. calculation step 0705 of FIG. 7 is used for .mu. and moreover the value obtained in the regression analysis step 0605 of FIG. 6 is used for .sigma..sup.2. In the temperature estimation step 0804, temperature S[C] of the C-th day is estimated depending on the [Expression 19]. S[C]=(1-.beta.).THETA.'[C]+.beta.S[C-1]+U [Expression 19] In the estimation period end determination step 0805, whether the value of the estimation day counter C is equal to the number of days of estimation period N' or not is determined. When these are not equal, the process shifts to the estimation day addition step 0806 and when these are equal, the process shifts to the estimation temperature time-series accumulation step 0807, respectively. In the estimation day addition step 0806, the number of days of estimation C is counted up. In the estimation temperature time-series accumulation step 0807, the estimation temperature time-series sequence S[l](l=0, . . . , N') is copied to the two-dimensional array for data accumulation R[j][l] (j=1, . . . , L; l=1, . . . , N') to become R[P][l]=S[l](l=1, . . . , N'). In the simulation end determination step 0808, whether the simulation number P is equal to the number of times of simulation L or not is determined. When these are not equal, the process shifts to the simulation number addition step 0809 and when these are equal, the process comes to the end and the process then shifts to the meteorological quantities estimation result output step 0204, respectively. Next, the meteorological quantities estimation result output step 0204 of FIG. 2 will be described with reference to FIG. 2 and FIG. 8. The meteorological quantities estimation result output step 0204 of FIG. 2 outputs the meteorological quantities estimation result 0214 of FIG. 2 based on the two-dimensional array R[J][l] (J=1, . . . , L; l=1, . . . ,N') of data accumulated in the estimation temperature time-series accumulation step 0807 of FIG. 8. Next, an example of the meteorological quantities estimation result 0214 of FIG. 2 will be described with reference to FIG. 9. The meteorological quantities estimation result 0214 of this embodiment is composed of date 091, simulation number 092 and estimation temperature 093 of each date of each simulation number. In FIG. 9, for example, the estimation temperature of a day of the date July 1, of the simulation number is 25.3[.degree. C.]. Next, an embodiment (first embodiment) of the meteorological quantities estimation system of the present invention will be described with reference to FIG. 11 to FIG. 14. The meteorological quantities estimation system of this embodiment is configured, as illustrated in FIG. 11, with a display 111, a CPU (Central Processing Unit) 112, a ROM (Read Only Memory) 113 storing a system activation program 1131, a keyboard 114, a mouse 115, a main memory 116, an operating system 1171, an external storage unit 117 storing a meteorological quantities estimation program 1172 and historical data of meteorological quantities 0211 which are connected with a communication bus 118. First, when a hardware of the meteorological quantities estimation system is activated, the operating system 1171 is loaded to the main memory 116 as illustrated in FIG. 12. Moreover, when an application of the meteorological quantities estimation system is activated, the operating system 1171 and meteorological quantities estimation program 1172 are loaded to the main memory 116 as illustrated in FIG. 13. In addition, when estimation of meteorological quantities is executed, the meteorological quantities estimation program 1171 estimates, as illustrated in FIG. 14, the meteorological quantities by reading the historical data 0211 of meteorological quantities and the long-range weather forecast data 0212 and estimation condition 0213 inputted through the display 111, keyboard 114 and mouse 115 and then outputs the meteorological quantities estimation result 0214 to the external storage unit 117. Next, an example of the meteorological quantities estimation program 1172 of FIG. 11 to FIG. 14 will be described with reference to FIG. 15. The meteorological quantities estimation program of this embodiment is composed of a data acquisition module 151, a meteorological time-series model creation module 152, a meteorological quantities simulation module 153 and a meteorological quantities estimation result output module 154. In addition, the meteorological time-series model creation module 152 is composed of a meteorological time-series model parameter estimation module 1521 and a meteorological time-series model parameter adjustment module 1522. The data acquisition module 151, meteorological time-series model creation module 152, meteorological quantities simulation module 153, meteorological quantities estimation result output module 154, meteorological quantities model parameter estimation module 1521 and meteorological time-series model parameter adjustment module 1522 are respectively program modules to execute the processes conforming to the data acquisition step 0201, meteorological time-series model creation step 0202, meteorological quantities simulation step 0203, meteorological quantities estimation result output step 0204, meteorological time-series model parameter estimation step 02021 and the meteorological time-series model parameter adjustment step 02022 illustrated in FIG. 12. An embodiment (second embodiment) of the meteorological quantities estimation system of the second embodiment will be described with reference to FIG. 16 to FIG. 19. The second embodiment is the embodiment for transmitting and receiving information via the communication network. The meteorological quantities estimation system of the second embodiment comprises, as illustrated in FIG. 16, a client 161, a communication network 162, a server 163, a communication line 164 connecting the client 161 and communication network 162 and a communication line 165 connecting the server 163 and communication network 162. Moreover, the client 161 is configured with a display 1611, a CPU 1612, a ROM 1613 storing a system activation program 16131, a keyboard 1614, a mouse 1615, a main memory 1616, an operating system 16171, an external storage unit 1617 storing the communication module 16172 and a communication unit 1618 which are connected through a communication bus 1619. Here, the communication module 16172 is a program module for exchanging information of the communication network 162 with the client 161 via the communication unit 1618 and communication line 164. Moreover, the server 163 is configured with a CPU 1631, a ROM 1632 storing a system activation program 16321, a communication unit 1633, a main memory 1634, an operating system 16351, a meteorological quantities estimation program 16352, a communication module 16353 and an external storage unit 1635 storing the historical data of meteorological quantities 0211 which are connected via a communication bus 1636. Here, the communication module 16353 is the program module for exchanging information of communication network 162 with the server 163 via the communication unit 1633 and communication line 165. When the client and hardware of server of the meteorological quantities estimation system are activated, the operating system 16171 is loaded to the main memory 1616 on the side of client 161, while the operating system 16351 is loaded to the main memory 1634 on the side of server 163. Moreover, when applications of client and server in the meteorological quantities estimation system are activated, the communication module 16172 is loaded to the main memory 1616 on the side of client 161 as illustrated in FIG. 18, while the meteorological quantities estimation program 16352 and communication module 16353 are also loaded to the main memory 1634 on the side of server 163. Moreover, when the meteorological quantities are estimated, as in FIG. 19, the meteorological quantities estimation program 16352 executes estimation of meteorological quantities by reading the historical data of meteorological quantities 0211 and the long-range weather forecast data 0212 and estimation condition 0213 inputted via the display 1611, keyboard 1614 and mouse 1615 of the client and then outputs the meteorological quantities estimation result 0214 to the external storage unit 1617 of the client. The meteorological quantities estimation program 16352 is configured in the same format as that of FIG. 15. Next, an example of the input display image of the long-range weather forecast data in the meteorological quantities estimation system of the first and second embodiments of the present invention will be described with reference to FIG. 20. The input display image of long-range weather forecast data as an example is composed of a long-range weather forecast data input pull-down menu 201 and a probability distribution input pull-down menu 202. A result of input of this input display image is then inputted to the meteorological quantities estimation program as the long-range weather forecast data 0212. The information to be displayed in the input display image of the long-range weather forecast data is never limited to this information. For example, in the second embodiment, the data for input display image is transmitted from the sever 163 and formed as the input display image by a browser program (not illustrated) of the client 161. Next, an example of the input display image of the estimation condition of the meteorological quantities estimation system in the first and second embodiments of the present invention will be described with reference to FIG. 21. An input display image of the estimation condition indicated as an example is composed of a estimation period input pull-down menu 211, an estimation point input 212 and a simulation number input pull-down menu 213. The result inputted through this input display image is then inputted to the meteorological quantities estimation program as the estimation condition 0213. The information for displaying the estimation condition to the input display image is not limited to such information. For instance, the data for displaying the input display image is transmitted, for example, from the server 163 and is formed as an input display image with a browser program (not illustrated) of the client 161. The present invention described above is never limited to the embodiments described above and allows various changes and modifications. Namely, the Dischel model described above is only an example and the long-range weather forecast data can also be fetched for models in the same manner even when various meteorological time-series models are used. Moreover, for example, the long-term forecast data for one month can provide probability distribution of average temperature during the period and amount of accumulated rainfall during the period for every first week, second week and third to fourth weeks of the coming month, in addition to the probability distribution of average temperature during the period and amount of accumulated rainfall during the period for one month in future, but the present invention can easily expand the embodiments thereof to fetch the information described above. For example, it is enough when the parameter adjustment process is executed for every first, second, third and fourth weeks and the parameters corresponding to each section for the simulation are used. Moreover, the second embodiment, for example, can also provide, via the communication network, the charged services to the client where the client is charged and also the charged services where the client is not charged. According to the present invention, since the meteorological quantities such as temperature and amount of rainfall of each day in future can be estimated not only by the historical data in the past but also by reflection of the long-range weather forecast data, the influence on the economic activities can be evaluated more accurately. System And Method Implementation Portions of the present invention may be conveniently implemented using a conventional general purpose or a specialized digital computer or microprocessor programmed according to the teachings of the present disclosure, as will be apparent to those skilled in the computer art. Appropriate software coding can readily be prepared by skilled programmers based on the teachings of the present disclosure, as will be apparent to those skilled in the software art. The invention may also be implemented by the preparation of application specific integrated circuits or by interconnecting an appropriate network of conventional component circuits, as will be readily apparent to those skilled in the art. The present invention includes a computer program product which is a storage medium (media) having instructions stored thereon/in which can be used to control, or cause, a computer to perform any of the processes of the present invention. The storage medium can include, but is not limited to, any type of disk including floppy disks, mini disks (MD's), optical disks, DVD, CD-ROMS, micro-drive, and magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices (including flash cards), magnetic or optical cards, nanosystems (including molecular memory ICs), RAID devices, remote data storage/archive/warehousing, or any type of media or device suitable for storing instructions and/or data. Stored on any one of the computer readable medium (media), the present invention includes software for controlling both the hardware of the general purpose/specialized computer or microprocessor, and for enabling the computer or microprocessor to interact with a human user or other mechanism utilizing the results of the present invention. Such software may include, but is not limited to, device drivers, operating systems, and user applications. Ultimately, such computer readable media further includes software for performing the present invention, as described above. Included in the programming (software) of the general/specialized computer or microprocessor are software modules for implementing the teachings of the present invention, including, but not limited to, acquiring estimation conditions for meteorological quantities including historical data of meteorological quantities such as temperature and amount of rainfall observed in the past, acquiring an estimation period for estimating meteorological quantities at an estimation point, acquiring a number of times for a simulation, acquiring long-range weather forecast data provided by the Meteorological Agency for the meteorological quantities during the estimation period at the estimation point, creating a meteorological time-series model for the meteorological quantities during the estimation period at the estimation point based on acquired historical data of meteorological quantities and acquired long-range weather forecast data, conducting for the number of times the simulation using the meteorological time-series model, and outputting a meteorological quantities estimation result based on the simulation using the meteorological time-series model, according to processes of the present invention. In the foregoing specification, the invention has been described with reference to specific embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
|
Same subclass Same class Consider this |
||||||||||