|
|
|
Çѱ¹¼öÀÚ¿øÇÐȸ / v.44, no.2, 2011³â, pp.135-144
|
Á÷Á¢ÀûÀÎ ¸Å°³º¯¼ö ÃßÁ¤¹æ¹ýÀ» ÀÌ¿ëÇÑ »õ·Î¿î ¼öÁ¤µÈ Neyman-Scott ±¸ÇüÆÞ½º¸ðÇü °³¹ß ¿¬±¸
( A Study of New Modified Neyman-Scott Rectangular Pulse Model Development Using Direct Parameter Estimation ) |
| ½ÅÁÖ¿µ;ÁÖ°æ¿ø;ÇãÁØÇà; ¿¬¼¼´ëÇб³ »ê¾÷±â¼ú¿¬±¸¼Ò;¿¬¼¼´ëÇб³ ÀϹݴëÇпø Åä¸ñȯ°æ°øÇаú;¿¬¼¼´ëÇб³ »çȸȯ°æ½Ã½ºÅÛ°øÇкΠÅä¸ñȯ°æ°øÇÐ;
|
|
|
 |
|
| |
| ÃÊ ·Ï |
| Á÷Á¢ÀûÀÎ ¸Å°³º¯¼ö ÃßÁ¤¹æ¹ýÀÇ ´Ù¾çÇÑ Neyman-Scott ±¸ÇüÆÞ½º¸ðÇü(NSRPM) ±â¹Ý ¸ðÇü¿¡ ´ëÇÑ Àû¿ë¼º °ËÅä¿Í Á¤±ÔºÐÆ÷¸¦ ÀÌ¿ëÇÑ »õ·Î¿î NSRPM(NMNSRPM)ÀÇ °³¹ß ¿¬±¸¸¦ ¼öÇàÇÏ¿´´Ù. ±â»óû ¼¿ï À¯ÀΰüÃø¼Ò¿¡¼ Á¦°øÇÏ´Â 49³âÀÇ °üÃø °¼öÀڷḦ »ç¿ëÇÏ¿© ¸Å°³º¯¼ö¸¦ ÃßÁ¤ÇÏ¿´À¸¸ç, ÃßÁ¤µÈ ¸Å°³º¯¼öµéÀÇ Á¤È®µµ¸¦ ÆÇ´ÜÇϰíÀÚ »ý¼ºµÈ °¼öÀÚ·áÀÇ Åë°è°ª, À¯°¼öÀÏ ºñÀ², °¼öºÐÆ÷¸¦ ºñ±³ÇÏ¿´´Ù. Åë°è°ªÀ» ºñ±³Çغ» °á°ú NSRPM°ú ¼öÁ¤ NSRPM(MNSRPM)Àº 7-9¿ùÀÇ °¼öÀÚ·á Åë°è°ªÀÇ Àý´ë»ó´ë¿ÀÂ÷°¡ Ä¿Áö´Â °ÍÀ» È®ÀÎÇÒ ¼ö ÀÖ¾úÀ¸¸ç, Àý´ë»ó´ë¿ÀÂ÷°¡ 10.11%·Î NMNSRPMÀÌ °¼öÀÚ·áÀÇ Åë°è°ª¸¦ °¡Àå Àß ¸ðÀÇÇÑ °ÍÀ¸·Î ³ªÅ¸³µ´Ù. À¯°¼öÀÏ ºñÀ²À» ºñ±³Çغ» °á°ú MNSRPMÀÇ Àý´ë»ó´ë¿ÀÂ÷ Æò±ÕÀÌ 16.35%·Î °¡Àå ÀÛÀº Àý´ë»ó´ë¿ÀÂ÷ °ªÀ» º¸¿´°í ±×·¡ÇÁ¸¦ ÀÌ¿ëÇÑ µµ½ÃÀûÀÎ ºÐ¼®¹ýÀ» ÅëÇÏ¿© ¼¼ ¸ðÇüÀÌ À¯°¼öÀÏ ºñÀ²À» °ú¼ÒÃßÁ¤ÇÏ´Â °ÍÀ» È®ÀÎÇÏ¿´´Ù. °¼öºÐÆ÷¸¦ ºñ±³Çغ» °á°ú ¼¼ ¸ðÇüÀÌ ¾à 2% ³»¿ÜÀÇ Àý´ë»ó´ë¿ÀÂ÷¸¦ º¸¿© ¼¼ ¸ðÇü ¸ðµÎ °¼öºÐÆ÷¸¦ Àß ¸ðÀÇÇÏ´Â °ÍÀ» È®ÀÎ ÇÏ¿´´Ù. Á÷Á¢ÀûÀÎ ¸Å°³º¯¼ö ÃßÁ¤¹æ¹ýÀ¸·Î NSRPM, MNSRPM, NMNSRPMÀÇ ¸Å°³º¯¼ö¸¦ ÃßÁ¤ ÇÒ ¼ö ÀÖ´Â °ÍÀ» È®ÀÎ ÇÏ¿´À¸¸ç, Á÷Á¢ÀûÀÎ ¸Å°³º¯¼ö ÃßÁ¤¹æ¹ýÀÌ NSRPM »Ó¸¸ ¾Æ´Ï¶ó À̸¦ ±â¹ÝÀ¸·Î ÇÑ ´Ù¸¥ ¸ðÇüµéÀÇ ¸Å°³º¯¼öµµ ÃßÁ¤ÇÒ ¼ö ÀÖ´Ù´Â °ÍÀ» È®ÀÎÇÏ¿´´Ù. NMNSRPMÀÇ ¸ðÀÇ Á¤È®µµ¸¦ ºñ±³ÇÑ °á°ú Á÷Á¢ÀûÀÎ ¸Å°³º¯¼ö ÃßÁ¤¹æ¹ýÀ» ÅëÇÑ NSRPM ±â¹ÝÀÇ »õ·Î¿î ¸ðÇü¿¡ ´ëÇÑ °³¹ßÀÌ °¡´ÉÇÏ´Ù´Â °ÍÀ» È®ÀÎÇÒ ¼ö ÀÖ¾úÀ¸¸ç, ¸ðÇüÀÇ ¼º´ÉÀÌ ±âÁ¸ ¸ðÇüµé°ú ºñ½ÁÇÑ ¼öÁØÀÓÀ» È®ÀÎÇÏ¿´´Ù. |
|
| Direct parameter estimation method is verified with various models based on Neyman-Scott rectangular pulse model (NSRPM). Also, newly modified NSRPM (NMSRPM) that uses normal distribution is developed. Precipitation data observed by Korea Meteorological Administration (KMA) for 47 years is applied for parameter estimation. For model performance verification, we used statistics, wet ratio and precipitation accumulate distribution of precipitation generated. The comparison of statistics indicates that absolute relative error (ARE)s of the results from NSRPM and modified NSRPM (MNSRPM) are increasing on July, August, and September and ARE of NMNSRPM shows 10.11% that is the smallest ARE among the three models. NMNSRPM simulates the characteristics of precipitation statistics well. By comparing the wet ratio, MNSRPM shows the smallest ARE that is 16.35% and by using the graphical analysis, we found that these three models underestimate the wet ratio. The three models show about 2% of ARE of precipitation accumulate probability. Those results show that the three models simulate precipitation accumulate probability well. As the results, it is found that the parameters of NSRPM, MNSRPM and NMNSRPM are able to be estimated by the direct parameter estimation method. From the results listed above, we concluded that the direct parameter estimation is able to be applied to various models based on NSRPM. NMNSRPM shows good performance compared with developed model-NSRPM and MNSRPM and the models based on NSRPM can be developed by the direct parameter estimation method. |
| |
| Ű¿öµå |
| Á¡°úÁ¤;°¿ì¸ðÇü;Á÷Á¢ÀûÀÎ ¸Å°³º¯¼ö ÃßÁ¤;NSRPM;MNSRPM;NMNSRPM;point process;rainfall model;direct parameter estimation; |
| |
|
|
 |
|
Çѱ¹¼öÀÚ¿øÇÐȸ³í¹®Áý / v.44, no.2, 2011³â, pp.135-144
Çѱ¹¼öÀÚ¿øÇÐȸ
ISSN : 1226-6280
UCI : G100:I100-KOI(KISTI1.1003/JNL.JAKO201108863881019)
¾ð¾î : Çѱ¹¾î |
|
| ³í¹® Á¦°ø : KISTI Çѱ¹°úÇбâ¼úÁ¤º¸¿¬±¸¿ø |
|
|
|
|
|
|