4. Results and discussion
During the examination of El-Ni~no=Southern
Oscillations (ENSO) and its relationship with
Indian monsoon, it was found that during ENSO
years (1983; 1987); the monsoon rainfall over
most part of the region was below normal, particularly
in the west area (Krishnamurti et al,
1989). It is known that within many places, the
proportion of rain comprises about 40–50% during
1–5 days and sometimes, over 100% of the
total mean seasonal rainfall. Such an occurrence
of rainfall extremes produces largest flooding
during monsoon season. It is also possible that
ENSO conditions may also affect the rainfall
pattern.
EOF analysis of the rainfall and OLR data for
68 stations has been carried out and the results
are presented in subsequent table and figures.
Table 1 shows the percentage of variance explained
by various principal components for rainfall
and OLR data in a decreasing order. As can
be seen, principal components up to 5 levels explain
the variance of 92.20% for rainfall and the
accompanying variance of 99.50% for OLR data.
This table also shows that OLR has higher spatial
coherence than rainfall. It is clear from Table 1
that each principle component has a pivotal role
in accounting for the corresponding values of
rainfall variance.
Figure 2a, b shows the first principal component
explaining the percentage of total variance.
It has been observed that for rainfall and OLR
data, the first principal component explains
55.50% and 58.00% of total variance respectively
and exhibits factor of one positive value
throughout the sub-continent. Further, it has also
been noticed from rainfall pattern (Fig. 2a) that
the large positive values are concentrated over
northeast as well as southwest coast of India.
Such a pattern has been explained by Shukla
(1987) due to Southern Oscillation while Bedi
and Bindra (1980) attributed it to monsoon depression
in the head Bay of Bengal, and midtropospheric
low over westcoast of India.
The impact in rainfall over Northeast and
Northwest India reveals an interesting phenomenon.
Due to depression in the Bay of Bengal or
southward movement of monsoon trough, the
rainfall amount increases over Northeast India.
High values of first principal components around
the westcoast and over Northeast India could
possibly be explained by the role played by topography
in bringing rainfall to these areas, when
the monsoon current passes over them.
However, the OLR pattern (Fig. 2b) shows that
the area of large positive values are concentrated
over Northwest India and lower values over
South India and elsewhere: Indian Ocean. The
Northwest India is covered with seasonal heatlow
and South Indian Ocean with seasonal
high-pressure area. The intense heat low, with
little or no clouds would contribute to higher values
of OLR. Correspondingly, Indian Ocean High
that is associated with clouds, will contribute to
lesser OLR magnitude. It appears that major reason
for the monsoon variability is the intensity
and the fluctuations in the two major semi-permanent
seasonal systems.
Figure 3a shows second principal component
of rainfall pattern explaining 15.80% of the variance
besides exhibiting factor of one positive
value throughout northeast side and also in
Konkan and Karnataka Coast, whereas most part
of the country shows the lower rainfall pattern.
This could be possible due to movement of monsoon
lows=depressions in quick succession in
the North Bay of Bengal, which also favours
the development of a mid troposhperic low over
west India. Figure 3b shows the second principal
components of OLR pattern explaining 18.00%
of the total variance exhibiting factor of a large
variation along 25 N over the north India region
optic systems such as monsoon depression=
lows over the north bay and trough=vortices off
westcoast in the Arabian sea.
Figure 4a, b depicts the third principal component
explaining about 10.40% of the total
variance for rainfall pattern and the third principal
components explaining 12.00% of the total
variance for OLR pattern. It has been noticed that
certain positive and negative value are concentrated
within certain zone (Fig. 4a). It could be
with positive values over Bengal and adjoining
region and negative value over Northwest
India.
The second region of variation, though weaker
as compared to the northern is located along
12 N and is reverse in direction, i.e., higher
values over westcoast and lower values over
eastcoast. The probable reason could be the syn
possible that the monsoon trough is in opposition
to the Himalayan foothills and the trough lays
oscillating north–south. The same may be developing
along it, while from OLR pattern (Fig. 4b)
negative values along the region of monsoon
trough line and positive value east–west are
oriented 12–13 N. One possible reason could
be variation from active to break monsoon and
back.
Figure 5a, b shows the fourth principal component
for rainfall and OLR pattern explaining
7.50% of the variance for rainfall, while the
fourth principal components explaining 8.00%
of the variance for OLR pattern. It has been
found from Fig. 5a that positive values are concentrated
only in north-east side, while as major
part of the country shows negative value. This
could be possible due to movement of monsoon
trough to the north, thus giving a break in the
monsoon rainfall. The deeper depression in the
northwest direction form in the head Bay of
Bengal looks to be a reason for the observed pattern,
while from Fig. 5b, large positive values
occur over Madhya Pradesh, Orissa, Karnatka
and Tamilnadu. This is in direct comparison
to the negative values demonstrated in case
of Maharashtra, Gujarat, and Rajasthan. Such
type of variations are possible at the time of
withdrawal of monsoon, such as September
month.