Geoscience Reference
In-Depth Information
the BS values exhibited by all the four experiments for various threshold values
with respect to the TRMM observations. All the four experiments reveal a BS value
greater than one for all the thresholds, indicating that all the four model experiments
are over estimating the precipitation features at all the thresholds. As expected, the
MODIS experiment has the lowest BS values above one (i.e., lowest overestimation)
at all the thresholds which indicates that the results of the MODIS experiment is
relatively closer to the observations. The CTRL and the SSM/I run show higher BS
values which indicate that both the above experiments have the highest degree of
overestimation of the intensity of the rainfall.
Improvement Parameter (
˜
)
“Improvement parameter” for the 24 h accumulated precipitation valid at 12 UTC of
21 September 2006 and 22 September 2006 are calculated. Improvement parameter
for any variable can be defined as,
˜
D j
.
observation control
/
j j
.
observation experiment
/
j
(26.4)
A positive value of “improvement parameter” is a clear indication of the positive
impact of assimilation of observation and vice versa. The results of the MODIS
experiment shows both positive and negative values of “improvement parameter”
over the domain of study (Fig.
26.10
). The maximum positive of
can be seen
on the day-one of the forecast having values up to about 150. Negative impact of
assimilation can be seen over the north of the domain. For the second day of forecast,
the positive impact is more prominent over the Bay of Bengal region than over
the land for the MODIS experiment. The ATOVS experiment also shows the same
patterns of spatial distribution of “improvement parameter” as that of the MODIS
experiment with highest positive values over the land. However, the negative impact
is less over the land for the ATOVS experiment when compared to the MODIS
experiment on day one of the forecast. In contrast to the first day, the second
day exhibits more negative
˜
over the land with values reaching up to -20. Also
the positive impact over the Bay of Bengal region is not as prominent as that of
the MODIS run. Hence, it can be concluded that despite the improvements in the
ATOVS experiment improvement parameter on the first day of the forecast, the same
decreased on the second day of the forecast. Undesirably large analysis increment
values generated due to the assimilation of the ATOVS temperature and humidity
profiles may be one of the reasons for the disappointing results. For the SSM/I
run, the day one forecast shows smaller extent of positive values of “improvement
parameter” over the domain. Furthermore, even though the second day forecast of
the SSM/I run has produced positive
˜
values over the oceans, the positive impact
is found to be lower over the land regions.
˜
Search WWH ::
Custom Search