14:00 〜 14:15
[MIS01-12] Quantitative Evaluation of Winter Precipitation in Mountainous Areas on the Japan Sea Side of Japan
- Validation based on snow weight observations -
キーワード：降水、雪、APHRODITE、山岳
The Japan Sea side of the country is a region with heavy snowfall, which is unique in the world. Heavy snowfalls can cause avalanches in mountainous areas and traffic disruption on the plains, and in some cases, loss of life. On the other hand, mountain snowfall is extremely important for the livelihood of the people in this region and its basin, and therefore, accurate winter precipitation data is required. However, it is difficult to accurately determine precipitation amounts in mountainous areas because of the lack of observation sites for the Japan Meteorological Agency's AMeDAS (the Japan Meteorological Agency's regional meteorological observatory system).
The purpose of this study is to estimate winter precipitation more precisely by verifying the following three types of precipitation grid data for mountainous areas using snow weight data from the National Research Institute for Earth Science and Disaster Prevention's Snow Weather Observation Network (SW-Net) at nine sites in high-elevation areas. The SW-Net data was used to convert the difference in weight from the previous day into daily precipitation values. APHRO_JP (Kamiguchi et al., 2010) interpolates the ratio to climatology. In this study, NHRCM climatological precipitation are used. The winter daily precipitation data from 2009 to 2012 were used for the evaluation.
1) JMA radar AMeDAS analysis (RA)
2) APHRO_JP
3) APHRO_JP with wind correction
The results of the monthly precipitation verification using snow weight observations showed that the order of error was 3) > 2) > 1), and although RA has more rain gauge locations than APHRO_JP, we consider that the quantitative performance of 3) and 2) is superior because the climate value correction was not performed. The results of the site-by-site verification showed that the corrected APHRO_JP and snow weight data generally agreed at Uonuma and Okutadami in Niigata Prefecture, but the corrected APHRO_JP underestimated the snow weight data significantly at Mt.Iwaki. One of the reasons for this may be that the resolution (0.05°) did not fully express the increase in precipitation corresponding to the elevation gradient in the climatological values. It is speculated that it is caused by the relationship with surrounding stations and wind direction, but further investigation is needed. The results showed that 3) > SW-Net when the bias of 3) with SW-Net was more than 20%, while 3) < SW-Net when the temperature was high and the wind was light, and 3) > SW-Net when the temperature was low and the wind was strong. The SW-Net can be regarded as the true value in the case of snow, but it may runoff in the case of rain and underestimate the true value. When the temperature is high, the correction 3) > SW-Net, it is possible that SW-Net underestimates the true value rather than APHRO_JP overestimates it. When the temperature is low and the wind is strong, it is possible that the correction for the capture loss is insufficient.
Next, SW-Net was added to 2) and 3) to examine precipitation distribution from plains to mountains along the prevailing wind direction in Niigata Prefecture (Nagaoka - Tochiotashiro - Uonuma - Okutadami).
4) 2) + SW-Net
5) 3) + SW-Net
As a result, the precipitation grid data in the mountainous areas in 4) and 5) were significantly increased and the profiles were improved. Quantitatively, the results were 5) > 4) > 3) > 2), indicating that the effect of the addition of SW-Net on quantification was much larger than the effect of the capture rate correction 5) - 4) and 3) - 2).
The purpose of this study is to estimate winter precipitation more precisely by verifying the following three types of precipitation grid data for mountainous areas using snow weight data from the National Research Institute for Earth Science and Disaster Prevention's Snow Weather Observation Network (SW-Net) at nine sites in high-elevation areas. The SW-Net data was used to convert the difference in weight from the previous day into daily precipitation values. APHRO_JP (Kamiguchi et al., 2010) interpolates the ratio to climatology. In this study, NHRCM climatological precipitation are used. The winter daily precipitation data from 2009 to 2012 were used for the evaluation.
1) JMA radar AMeDAS analysis (RA)
2) APHRO_JP
3) APHRO_JP with wind correction
The results of the monthly precipitation verification using snow weight observations showed that the order of error was 3) > 2) > 1), and although RA has more rain gauge locations than APHRO_JP, we consider that the quantitative performance of 3) and 2) is superior because the climate value correction was not performed. The results of the site-by-site verification showed that the corrected APHRO_JP and snow weight data generally agreed at Uonuma and Okutadami in Niigata Prefecture, but the corrected APHRO_JP underestimated the snow weight data significantly at Mt.Iwaki. One of the reasons for this may be that the resolution (0.05°) did not fully express the increase in precipitation corresponding to the elevation gradient in the climatological values. It is speculated that it is caused by the relationship with surrounding stations and wind direction, but further investigation is needed. The results showed that 3) > SW-Net when the bias of 3) with SW-Net was more than 20%, while 3) < SW-Net when the temperature was high and the wind was light, and 3) > SW-Net when the temperature was low and the wind was strong. The SW-Net can be regarded as the true value in the case of snow, but it may runoff in the case of rain and underestimate the true value. When the temperature is high, the correction 3) > SW-Net, it is possible that SW-Net underestimates the true value rather than APHRO_JP overestimates it. When the temperature is low and the wind is strong, it is possible that the correction for the capture loss is insufficient.
Next, SW-Net was added to 2) and 3) to examine precipitation distribution from plains to mountains along the prevailing wind direction in Niigata Prefecture (Nagaoka - Tochiotashiro - Uonuma - Okutadami).
4) 2) + SW-Net
5) 3) + SW-Net
As a result, the precipitation grid data in the mountainous areas in 4) and 5) were significantly increased and the profiles were improved. Quantitatively, the results were 5) > 4) > 3) > 2), indicating that the effect of the addition of SW-Net on quantification was much larger than the effect of the capture rate correction 5) - 4) and 3) - 2).