109
maximum elevation in western Himalaya is seen. This separation of southerly and
northerly winds creates a westerly trough over this region. Over the eastern
Himalaya, the meridional winds are different than those over the western parts.
Winds are mostly southerlies in the WRF model simulations north of 34°N at all the
levels. These southerly winds are also seen to be sandwiched between northerly
winds from 24°N to 35°N between 600 and 400 hPa. The southerlies at 500hpa over
35°N is about 2 m/s stronger in the model than that in ERA-I. Over the eastern
Himalaya, the model simulates stronger northerly winds at all levels north of
35°N. Therefore, the WRF model is able to simulate all the essential features of the
winter-time circulation over the Himalaya with more sharp features.
Interannual variability (IAV) of seasonal mean precipitation (mm/d) over the
Himalaya simulated by the WRF model is shown in Fig. 5a. It is seen that there is
considerable IAV over the western Himalaya with a narrow band extending up to
central and eastern Himalaya. Maximum IAV (> 3–4 mm/day) is noticed over the
western Himalaya including Karakoram region, Jammu and Kashmir and Himachal
Pradesh covering the Sutlej River basin. It may be noted that this is the source
region of the Indus River and its tributaries. As already noted, about 80% or more of
the precipitation falling in this region is snow. The accumulation of this snow feeds
the glaciers and the snowmelt help sustain the river flow in the next spring and summer seasons. Therefore, analysis and prediction of precipitation in the areas where
there is large IAV is important for the economy and lives of south Asia. A lot has
been talked about the Karakoram anomaly and how the glaciers in Karakoram are
either advancing or stagnant in recent years while glaciers in the neighboring regions
and world-wide have retreated. In order to examine the region that co-vary with the
precipitation over Karakoram, correlation analysis of the model simulated precipitation over the Karakoram (72°E–77°E and 35°N–37°N) with those over the
Himalaya have been carried out (Fig. 5b). It is found that statistically significant
correlation (significant at 95% from a Student’s T-test) exist over the Hindukush and
western Himalaya. The central and eastern Himalaya do not exhibit any significant
correlation with the precipitation over Karakoram. Some parts of the Gangetic
plains and eastern Himalaya have negative correlation with precipitation over
Karakoram, though not significant. In order to further study the interannual variability of precipitation, area averaged model simulated seasonal mean precipitation
over 70°E–79°E and 33°N–38°N have been examined and shown in Fig. 5c. It is
seen that there is an increasing trend in the model simulated precipitation from the
year 2000. Based on this time-series of precipitation, 3 years with more precipitation (1991, 2004 and 2008) and 3 years with less precipitation (1996, 1999 and
2000) have been chosen as excess and deficit years for further study.
Composite analysis of the model simulated T2 m and precipitation for these
excess and deficit years have been examined. The differences (excess-deficit) are
shown in Fig. 6. It is found that when the precipitation is more over the chosen
High-Resolution Dynamic Downscaling of Winter Climate over the Himalaya
maximum elevation in western Himalaya is seen. This separation of southerly and
northerly winds creates a westerly trough over this region. Over the eastern
Himalaya, the meridional winds are different than those over the western parts.
Winds are mostly southerlies in the WRF model simulations north of 34°N at all the
levels. These southerly winds are also seen to be sandwiched between northerly
winds from 24°N to 35°N between 600 and 400 hPa. The southerlies at 500hpa over
35°N is about 2 m/s stronger in the model than that in ERA-I. Over the eastern
Himalaya, the model simulates stronger northerly winds at all levels north of
35°N. Therefore, the WRF model is able to simulate all the essential features of the
winter-time circulation over the Himalaya with more sharp features.
Interannual variability (IAV) of seasonal mean precipitation (mm/d) over the
Himalaya simulated by the WRF model is shown in Fig. 5a. It is seen that there is
considerable IAV over the western Himalaya with a narrow band extending up to
central and eastern Himalaya. Maximum IAV (> 3–4 mm/day) is noticed over the
western Himalaya including Karakoram region, Jammu and Kashmir and Himachal
Pradesh covering the Sutlej River basin. It may be noted that this is the source
region of the Indus River and its tributaries. As already noted, about 80% or more of
the precipitation falling in this region is snow. The accumulation of this snow feeds
the glaciers and the snowmelt help sustain the river flow in the next spring and summer seasons. Therefore, analysis and prediction of precipitation in the areas where
there is large IAV is important for the economy and lives of south Asia. A lot has
been talked about the Karakoram anomaly and how the glaciers in Karakoram are
either advancing or stagnant in recent years while glaciers in the neighboring regions
and world-wide have retreated. In order to examine the region that co-vary with the
precipitation over Karakoram, correlation analysis of the model simulated precipitation over the Karakoram (72°E–77°E and 35°N–37°N) with those over the
Himalaya have been carried out (Fig. 5b). It is found that statistically significant
correlation (significant at 95% from a Student’s T-test) exist over the Hindukush and
western Himalaya. The central and eastern Himalaya do not exhibit any significant
correlation with the precipitation over Karakoram. Some parts of the Gangetic
plains and eastern Himalaya have negative correlation with precipitation over
Karakoram, though not significant. In order to further study the interannual variability of precipitation, area averaged model simulated seasonal mean precipitation
over 70°E–79°E and 33°N–38°N have been examined and shown in Fig. 5c. It is
seen that there is an increasing trend in the model simulated precipitation from the
year 2000. Based on this time-series of precipitation, 3 years with more precipitation (1991, 2004 and 2008) and 3 years with less precipitation (1996, 1999 and
2000) have been chosen as excess and deficit years for further study.
Composite analysis of the model simulated T2 m and precipitation for these
excess and deficit years have been examined. The differences (excess-deficit) are
shown in Fig. 6. It is found that when the precipitation is more over the chosen
High-Resolution Dynamic Downscaling of Winter Climate over the Himalaya