Meteorol Atmos Phys (2011) 113:55–66 DOI 10.1007/s00703-011-0140-1
ORIGINAL PAPER
A study on radiative properties of Indian summer monsoon clouds V. Sathiyamoorthy • Bipasha Paul Shukla P. K. Pal
•
Received: 7 October 2010 / Accepted: 22 April 2011 / Published online: 19 May 2011 Ó Springer-Verlag 2011
Abstract Possible causes behind the unusual cooling by summer monsoon clouds over India are investigated. Results suggest that the causes behind the cooling over the Bay of Bengal, India (BBI) and Arabian Sea (AS) within the Indian monsoon region are different. Over the BBI, clouds are tall. A unique upper tropospheric easterly jet stream exists over India during the summer monsoon season, which horizontally spreads the vertically growing deep convective clouds and thereby increases the cloud cover. Hence, more incoming solar radiation is reflected back to space, which leads to cooling. A radiative transfer study employing the Santa Barbara DISORT Atmospheric Radiative Transfer model supports this view. Over the Arabian Sea, clouds are shallow, and hence the upper tropospheric jet cannot affect them. Due to their proximity to the ground, Arabian Sea clouds exert less warming effect, but they exert a considerable cooling effect, which arises because of the high reflectivity of the clouds. Over the Equatorial Indian Ocean (EIO), where the monsoon clouds originate and propagate towards the monsoon trough region, both cooling and warming effects are nearly canceled out. The upper tropospheric jet is located hundreds of kilometers north of the EIO, and hence it does not disturb the deep convective clouds of the EIO. Therefore, they behave similarly to other deep convective clouds in the tropical belt.
Responsible editor: J. Fasullo. V. Sathiyamoorthy (&) B. P. Shukla P. K. Pal Atmospheric and Oceanic Sciences Group, Space Applications Centre (ISRO), SAC-Post, Ahmedabad 380015, India e-mail:
[email protected];
[email protected]
1 Introduction Clouds affect the radiation balance of the earth-atmosphere system by both reflecting a fraction of incoming solar radiation back to space (cooling the near surface) and blocking a fraction of the earth emitting longwave radiation from escaping to space (greenhouse warming). The net effect is either cooling or warming. Cloud macro-physical properties such as the amount of cloud cover, cloud top height, etc., and cloud microphysical properties such as cloud droplet size, liquid-ice phase, etc., play a key role in cloud-radiation interaction. For example, relatively shallow stratus clouds tend to cool the earth-atmosphere system (cooling dominates warming), whereas upper tropospheric thin cirrus clouds tend to warm the system (warming dominates cooling). Weare (1997) showed that greenhouse warming is strongly related to high cloud amount, and cooling is more related to high cloud amount and cloud optical depth. Radiation budget studies by Kiehl (1994) suggested that tropical deep convective clouds neither cool nor warm in an average sense, i.e., cooling due to shortwave reflection is nearly balanced by warming due to longwave absorption. A surprising exception is found over the Asian summer monsoon region (which includes the Indian monsoon region as well) where the monsoon deep convective clouds tend to cool the earth-atmosphere system. Rajeevan and Srinivasan (2000) suggested that the large amount of highlevel clouds with high optical depth present over this region possibly causes this cooling. Over the Indian monsoon region, clouds appear to have spatially inhomogeneous physical properties during the summer monsoon season of June to September. For example, windward and leeward sides of the Western Ghat mountain chain, semi-arid regions of West India, the
123
56
monsoon trough region, etc., are expected to have different cloud physical properties because of the large differences in vertical motion, moisture availability and other meteorological parameters. Hence, monsoon clouds with diverse cloud physical properties are expected to interact differently with incoming solar and outgoing terrestrial radiations. In this article an attempt has been made to understand the radiative properties of Indian summer monsoon clouds during the summer monsoon season of June to September and in particular the causes behind the unusual cooling by monsoon clouds.
2 Data and methodology We used broadband top-of-atmosphere earth radiation budget flux quantities from the Earth Radiation Budget Satellite (ERBS) for the analysis (Harrison et al. 1990). ERBS is a part of the three-satellite Earth Radiation Budget Experiment, which provided useful data from November 1984 to February 1990. ERBS monthly mean radiation flux data (S4-G) are available at 2.5° latitude 9 2.5° longitude resolution. Reported accuracy of these fluxes is ±10 Wm-2 (Kiehl 1994). To date, ERBS is one of the most successful earth radiation budget missions because of (1) its better diurnal sampling capability and (2) continuous data availability for more than 5 years. Therefore, we mainly depended on this data for this study. The cloud-radiative forcing terms are defined as follows: Shortwave cloud radiative forcing (SWCRF) ¼ Sðaclr aÞ where S is the monthly mean incoming solar flux at the top of the atmosphere, and a and aclr are the total and clear-sky, respectively, albedo of the earth-atmosphere system: Longwave cloud radiative forcing (LWCRF) ¼ Fclr F where F and Fclr are all-sky and clear-sky longwave fluxes, respectively, at the top of the atmosphere: Net cloud radiative forcing (NCRF) = SWCRF + LWCRF: Generally SWCRF is a negative quantity, and LWCRF is a positive quantity. In addition, earth radiation budget data from clouds and the earth’s radiant energy system (CERES) onboard the Tropical Rainfall Measuring Mission (Wielicki et al. 1998) were also used to validate the results. The monthly mean top-of-atmosphere radiative flux data available during January to August 1998 with a horizontal resolution of 2.5° latitude 9 2.5° longitude were considered for this purpose.
123
V. Sathiyamoorthy et al.
Monthly cloud data from the International Satellite Cloud Climatology Project (ISCCP), version D2, for the same 5-year period were used to study cloud physical properties (Rossow et al. 1996). The ISCCP-D2 cloud data were available globally at 2.5° latitude 9 2.5° longitude resolution. Cloaud amount is the fractional area covered by clouds as observed from satellites, cloud-top pressure is determined from the cloud-top temperature, and cloud optical depth (COD) represents the optical thickness of clouds at visible wavelengths. Monthly mean circulation parameters are obtained from the National Centers for Environmental Prediction-National Center for Atmospheric Research (NCEP-NCAR) reanalysis project (Kalnay et al. 1996).
3 Results 3.1 Physical and radiative properties of the Indian monsoon clouds The 5-year (1985–1989) mean total cloud cover amount, cloud top pressure, COD and top-of-atmosphere cloud radiative forcing components (SWCRF, LWCRF and NCRF) averaged during the summer monsoon season are shown in Fig. 1a–f. Cloud cover was more than 70% over the entire Indian monsoon region (IMR, 55°E–100°E; 5°S– 30°N) except in the northwest corner. Cloud cover was highest (C90%) over the North Bay of Bengal. Clouds were taller (\450 hPa) over the Bay of Bengal, the equatorial Indian Ocean and most of the Indian land region (Fig. 1b). Over the Arabian Sea, cloud tops decreased westward, reaching as low as 750 hPa off the Saudi Arabian coast. Figure 1c suggests that the COD is more than 4 over the central Arabian Sea, West Coast and Central India and North Bay of Bengal. The magnitude of SWCRF is larger ([90 Wm-2) over (1) the central and north Bay of Bengal and adjoining land regions and (2) the west coast of India, where the cloud cover, cloud top height and COD are higher. The magnitude of SWCRF is lowest over the less cloudy AfghanistanSaudi Arabian regions. LWCRF is highest ([75 Wm-2) over the North Bay of Bengal. The LWCRF contours closely match with the cloud top pressure contours. It is noted here that though the peak SWCRF and LWCRF values occur over the North Bay of Bengal, the magnitude of SWCRF is higher than the magnitude of LWCRF. As a result, NCRF is negative over the North Bay of Begnal. NCRF is also negative over the entire Arabian Sea, though the clouds are not deep. Hence, it is clear from Fig. 1 that although net cooling (NCRF is negative) is found over the Bay of Bengal and Arabian Sea, cloud physical properties
Study on radiative properties
(a)
57
(c)
(b)
(d)
(e)
(f)
Fig. 1 The 5-year (1985–1989) average cloud physical parameters: a total cloud cover amount (%), b cloud top pressure (hPa), c cloud optical depth; and cloud-radiative parameters: d SWCRF (Wm-2), e LWCRF (Wm-2) and f NCRF (Wm-2) averaged during the
monsoon months (June–September) over the Indian monsoon region. Blanks in d–e are due to data gaps arising from the lack of sufficient clear-sky scenes in satellite observations over persistently cloudy regions
are different over these two regions. Over the EIO, the cloud cover amount and cloud top heights are larger. Magnitudes of SWCRF and LWCRF are nearly the same, and hence the NCRF is near zero.
summer monsoon season of all 5 years of ERBS operation (1985–1989) are shown for AS, BBI and EIO boxes along with the whole IMR region (Fig. 3a–d). It is interesting to note that the AS, BBI and EIO regions show distinctively different associations between SWCRF and LWCRF, which are otherwise difficult to infer from the scatter plot of the IMR. The AS and BBI boxes show a general imbalance, whereas the EIO shows a near balance between SWCRF and LWCRF. Over the AS, the magnitude of LWCRF is mostly less than for SWCRF, suggesting a net cooling by clouds during the monsoon season. Values of SWCRF range from near zero to -130 Wm-2, whereas LWCRF values range only between near zero and 80 Wm-2. A large population of scatter points has low values of LWCRF, which suggests that the clouds over the Arabian Sea are shallow in nature. The average value of the ratio N (N = -SWCRF/LWCRF) is 1.78, suggesting a strong cooling by the clouds.
3.2 Association between SWCRF and LWCRF To understand the cloud-radiation interaction in a detailed way, three square grid boxes having the sizes of 10° latitude 9 10° longitude were chosen from the IMR located over the BBI (77.5°–87.5°E, 15°–25°N), AS (60°–70°E, 15°–25°N) and EIO (90°–100°E, 5°S–5°N), representative of regions with diverse cloud-physical and cloud-radiative properties (Fig. 2). Association between SWCRF and LWCRF over the three square boxes was examined along with that of the entire IMR. Scatter plots between the monthly mean LWCRF and SWCRF at individual grid points during the
123
58
Fig. 2 Map showing the study area. The three boxes considered for detailed analysis are also shown
Over the BBI, a near balance between LWCRF and SWCRF is found when the magnitude of SWCRF and LWCRF is between zero and 60 Wm-2. Beyond this range, LWCRF values are clustered between 60 Wm-2 and 90 Wm-2, whereas SWCRF values range between -80 and Fig. 3 Scatter plot between monthly mean LWCRF (Wm-2) and SWCRF (Wm-2) from the ERBS at individual grid points for a the Arabian Sea (60°–70°E, 15°–25°N), b the BBI (77.5°–87.5°E, 15°–25°N), c the Equatorial Indian Ocean (90°–100°E, 5°S–5°N) and d the Indian Monsoon Region (55°–100°E, 5°S–30°N) during the summer monsoon season (June–September) of 1985–1989. The total number of data points used in each plot and the ratio N are also provided
123
V. Sathiyamoorthy et al.
-160 Wm-2, suggesting an imbalance between SWCRF and LWCRF. This imbalance causes cooling over the BBI. Close examination of the LWCRF and SWCRF data suggests that lower magnitudes of LWCRF and SWCRF occur over the BBI mainly during June and September, whereas an imbalance occurs mainly during the peak summer monsoon months of July and August. The average value of the ratio N is 1.28. Although the BBI and AS experience a net cooling by clouds (negative NCRF), the main difference between the two regions is that the large population of scatter points has lower magnitudes of LWCRF and SWCRF over the AS than over the BBI. To conclude, though clouds over the AS are shallow and over the BBI are deep, both exert net cooling, which requires further investigations to ascertain the causes behind the cooling. Figure 1f shows low values of NCRF (* ±10 Wm-2) over the EIO. Both LWCRF and SWCRF appear to have nearly the same magnitudes, so that when they are added, near zero NCRF values are obtained. The value of N is closer to 1. This near cancellation is similar to what was observed over tropical deep convective belts by Kiehl (1994). Nearly closer values of LWCRF over the BBI (particularly in the peak monsoon season) and the EIO suggest that the cloud tops of these two regions have nearly the same altitude. However, the magnitude of the SWCRF is larger over the BBI than over the EIO, which makes the BBI different from the EIO and from other tropical deep convective regions.
Study on radiative properties
3.3 Possible causes behind the negative NCRF According to Rajeevan and Srinivasan (2000), the large amount of high cloud cover with high COD present over
Fig. 4 Scatter plot between monthly mean high cloud cover amount (%) from ISCCP and LWCRF (Wm-2, left column)/SWCRF (Wm-2, right column) from the ERBS at individual grid points over three selected boxes for the monsoon months (June–September) of
59
the IMR causes a negative NCRF. In this section we examine the influence of these two cloud physical parameters on cloud radiative forcing components over the three boxes. In Fig. 4, scatter plots between high cloud amount
1985–1989. The slope of the linear fit line, total number of data points used, correlation coefficient (R) and p value (for the t test of the slope = 0) are also provided
123
60
V. Sathiyamoorthy et al.
and cloud radiative forcing components (SWCRF and LWCRF) for the three boxes are shown for the monsoon months during 1985–1989. First, it is obvious that the high cloud cover amount is less (only 14% of the grids have high cloud cover of more than 30%) over the Arabian Sea than over the BBI and EIO boxes. As suggested by Rajeevan and Srinivasan (2000), both the BBI and EIO boxes exhibit a strong association between high cloud cover and SWCRF/LWCRF. The correlation coefficients are high and are statistically significant (Table 1). Over the AS, the LWCRF shows a good association with high cloud amounts, whereas the SWCRF shows a comparatively weaker association with high cloud cover. In Fig. 5, scatter plots between COD and SWCRF/LWCRF are shown for the three boxes during the monsoon seasons of the same 5-year period. Both SWCRF and LWCRF show good association with COD in all three boxes, including the AS box. In particular, the association between SWCRF and COD is better over the AS than over the other two boxes. 3.4 Negative NCRF over the BBI region: a radiative transfer study Sathiyamoorthy et al. (2004) demonstrated that the large amount of high cloud cover found over the IMR comes from horizontal spreading of vertically growing deep convective clouds by the unique upper tropospheric tropical easterly jet stream (TEJ) present over the IMR (Fig. 6). A ship-based and space-based cloud radar study over the Bay of Bengal also confirmed such horizontal spreading at the upper tropospheric levels (Zuidema and Mapes 2008). Though these horizontally spread clouds come under the high cloud category (cloud top \440 hPa), they have higher optical depth than the normal fair weather cirrus clouds (Rajeevan and Srinivasan 2000). Infrared imagery taken by the Kalpana-1 geostationary satellite over the IMR during a typical monsoon day in 2010 is shown in Fig. 7. Westward spreading of clouds is clearly seen in the imagery. Finer scale details like the individual cloud
elements are clearly seen over the EIO where the upper tropospheric winds are relatively weaker and spreading is less pronounced. But such finer details are missing over the other parts because of the spreading at the upper levels by the TEJ. This type of strong cloud spreading on a large spatial scale is unique for the IMR. In this section, the influence of the horizontally spread upper tropospheric cloud layer on top-of-atmosphere flux quantities is studied using an idealistic radiative transfer (RT) model simulation. The RT simulations for this study are carried out using the RT code of the Santa Barbara DISORT Atmospheric Radiative Transfer (SBDART, Appendix). The core of the TEJ appears to slope downward from north to south. Also, the axis of the TEJ is not stationary, but periodically propagates northward on intraseasonal (30–60 days) time scales (Sathiyamoorthy et al. 2007). During the active (break) monsoon phase, the core of the TEJ is close to 5°N (15°N). For the RT model study, we assume (1) the TEJ starts to spread the deep convective clouds at 8-km altitude as suggested by Sathiyamoorthy et al. (2004) and spreading extends up to 8.25 km, and (2) above and below no cloud is present. So we consider a thin layer (0.25 km) of uniformly distributed cloud between 8- and 8.25-km altitude for RT simulation. The effective radius, which is defined as the ratio of the second and third moments of radius distribution, is a measure of cloud droplet size. In the present case, cloud is assumed to be composed of ice particles having an effective radius of 25 lm. The optical depth is varied from 1 to 10, and the top of the atmosphere shortwave and longwave fluxes are simulated for the sea surface in clear and cloudy conditions in order to compute cloud radiative forcing components (Fig. 8). RT simulation suggests that: (1)
For low COD values, SWCRF is lower than LWCRF, suggesting a net warming. Optically thin clouds like fair weather cirrus allow solar radiation to penetrate to lower levels, and hence they exert less cooling by the albedo effect. Because of their cold temperatures,
Table 1 Slopes and correlations between high cloud cover/COD and CRF components over the three boxes considered for the study Box
Parameter
High cloud cover (%) -2
AS BBI EIO
Cloud optical depth -2
LWCRF (Wm-2)
SWCRF (Wm-2)
LWCRF (Wm )
SWCRF (Wm )
Slope
1.32a
-1.39a
7.82b
Correlation
0.78
-0.51
0.69
-0.86
Slope
0.95a
-1.69a
7.6b
-16.72b
Correlation
0.82
-0.82
Slope
0.97a
-1.25a
Correlation
0.76
-0.66
0.61 10.0b 0.76
-15.36b
-0.75 -16.29b -0.83
Data during the monsoon months of 1985–1989 are considered for the analysis. Correlations are significant at the 95% level of confidence Units:
a
Wm-2/%,
123
b
Wm-2/COD
Study on radiative properties
61
Fig. 5 Scatter plot between monthly mean cloud optical depth from ISCCP and LWCRF (Wm-2, left column)/SWCRF (Wm-2, right column) from the ERBS at individual grid points over three selected boxes for the monsoon months (June–September) of 1985–1989.
Slope of the linear fit line, total number of data points used, correlation coefficient (R) and p value (for the t test of the slope = 0) are also provided
they emit low radiation to space. The net effect of this optically thin upper level cloud is warming. For a COD of more than 2, SWCRF is more than LWCRF, suggesting a net cooling. This is what
happens over the BBI box. This type of cloud reflects more incoming solar radiation and allows less solar radiation to reach the surface as the COD increases. Hence, cooling (by the albedo effect) increases as the
(2)
123
62
V. Sathiyamoorthy et al.
Fig. 8 SBDART simulation of top of atmosphere SWCRF (Wm-2) and LWCRF (Wm-2) as a function of cloud optical depth
Fig. 6 The 5-year (1985–1989) average 100-hPa wind (ms-1) from the NCEP reanalysis averaged during the monsoon months of June to September. The dark shaded area is the jet core
Fig. 7 Infrared imagery of the Kalpana-1 geostationary satellite taken at 1,000 UTC (1530 Indian standard time) on 30 July 2010. Strong convective activity is seen all over the Indian monsoon region. Bright clouds (cold cloud tops) are spread by the strong easterlies so that they appear to dissipate on the western sides of the convective systems. Clouds in the southeastern corner do not show such spreading because of the weak upper tropospheric winds
123
COD increases. Since the LWCRF is a function of cloud top height, it does not change much with the increase in COD. The net result is cooling. The experiment was repeated by increasing the cloud depth from 0.25 to 0.5 km (i.e., beginning at 8 km and extending up to 8.5 km), and it was observed that the results did not vary because of this change. Therefore, the RT simulation confirmed that stratiform upper level clouds with a high cloud cover amount and COD generated by the horizontal spreading of the TEJ caused cooling as envisaged by Sathiyamoorthy et al. (2004). So the most likely cause for the cooling over the BBI is cloud spreading by the TEJ. It is to be mentioned here that climatologically the summer monsoon sets in over the southern tip of the Indian peninsula on 1 June every year, gradually moves northward and takes 1 month to cover all of India. The monsoon is fully established over India in July and August. India receives more rainfall in these 2 months, and hence they are referred to as the peak monsoon months. The monsoon begins to withdraw from northern parts of India from 1 September. As mentioned earlier, near cancellation of SWCRF and LWCRF is seen over the BBI region in June and September when the monsoon either begins to cover or begins to withdraw from India. During these 2 months, the TEJ is not fully established over India. During the peak monsoon months of July and August, the TEJ is fully established over India with its full strength. Hence, the TEJ can spread cloud tops more effectively in peak monsoon months and cool the region when compared to the other 2 months. This is the possible cause for the pronounced imbalance between the SWCRF and LWCRF over the BBI in the peak monsoon months and near balance in June and September.
Study on radiative properties
63
Fig. 9 Scatter plot between monthly mean 500-hPa vertical velocity (Pa s-1) from NCEP reanalysis and LWCRF (Wm-2, left column)/ SWCRF (Wm-2, right column) from ERBS at individual grid points
over three selected boxes for the monsoon months (June–September) of 1985–1989. The slope of the linear fit line, total number of data points used and p value (for the t test of the slope = 0) are also provided
3.5 Negative NCRF over the AS
tropospheric thermal inversion during the summer monsoon season (Narayanan and Rao 1981). Likely causes behind the inversion are: (1) cooling of sea surface temperatures by upwelling caused by low-level monsoon
In this section, we examine the causes behind the cooling of the AS clouds. The AS is characterized by lower
123
64
Fig. 10 Scatter plot between monthly mean 500-hPa vertical velocity (Pa s-1) from NCEP reanalysis and cloud optical depth from ISCCP at individual grid points over the AS box for the monsoon months (June–September) of 1985–1989
winds; (2) dry-hot air intrusion from the nearby Saudi Arabian desert; (3) descending motion from monsoon convection. Hence, a likely cause for the cooling is the suppression of cloud growth by inversion. Therefore, we examined the association between CRF components with large-scale vertical motion over the AS. In Fig. 9, scatter plots between 500-hPa vertical velocity and CRF components are shown for the AS as well as the other two boxes. Over the EIO, generally ascending motion prevails (negative pressure vertical velocity), and the vertical velocity ranges from -0.6 to 0 Pa s-1. Over the BBI, a wide range of vertical velocity was found from mild descending to strong ascending motions. Both the SWCRF and LWCRF over the EIO and BBI showed good association with 500-hPa vertical velocity. In other words, the higher the ascending motion, the taller the clouds will be and the more the LWCRF will be. Similarly, higher ascending motion will generate taller clouds and more SWCRF because of the increased COD. Over the AS, a peculiar type of association is found between the vertical velocity and SWCRF/ LWCRF. First, the SWCRF is constant for the vertical velocity range of 0.06 to 0.03 Pa s-1, and acquires a value anywhere between 0 and -110 Wm-2 when the vertical velocity is around 0.02 Pa s-1. For the remaining vertical velocity range, the SWCRF increases as the strength of ascending motion increases, which is similar to what is found over the BBI and EIO regions. Over the AS, a large population of scatter points has a positive vertical velocity (descending motion) as mentioned by Narayanan and Rao (1981). We suggest that during strong descending motion, clouds cannot form, and the SWCRF will be near zero because of the negligible difference between the clear-sky and all-sky reflected SW fluxes. When mild descending
123
V. Sathiyamoorthy et al.
motion prevails, clouds may possibly form, but have suppressed growth. But instead of remaining close to zero, the SWCRF varies anywhere between 0 and -110 Wm-2. We speculate that the large amount of mineral dust aerosols brought by the low-level monsoon jet stream from nearby Somalia and Saudi Arabian regions to the AS may increase the COD by increasing the cloud droplet concentration. These aerosols act as cloud condensation nuclei. Increased cloud droplets will reflect more incoming solar radiation back to space and cause cooling (Twomey effect). The amount of aerosols pumped to the AS may depend on the direction of wind flow (more aerosols will be pumped if the flow is from the desert to the AS) and wind speed. Hence, SWCRF is highly variable though vertical velocity remains constant. LWCRF also shows a similar but inverted association because of the difference in the sign. In Fig. 10, the scatter plot between vertical velocity and COD is shown for the AS. As expected, around 0.2 Pas-1, the COD varies widely, thus supporting the above mechanism. A detailed analysis of the role of aerosols on top-of-atmosphere flux quantities over the Arabian Sea is required to validate this hypothesis. 3.6 Near cancellation of SWCRF and LWCRF over the EIO The EIO is a typical example of a tropical deep convective region where SWCRF and LWCRF cancel out each other in an average sense as pointed out by Kiehl (1994). The clouds are tall and reflective in nature. The cooling (albedo effect) is nearly balanced by the warming (greenhouse effect). If the TEJ was present over this region, the cloud tops would be spread by the TEJ, and cloud-radiation interaction would be comparable to that of the BBI box.
4 Discussion and conclusions Using top-of-atmosphere cloud radiative forcing data from the ERBS, we examined the possible causes behind the unusual cooling by clouds over the IMR. Results suggest that the mechanisms behind the cooling over the BBI and AS regions are different. Over the BBI, clouds are tall, and the cloud tops are horizontally spread westward by the upper tropospheric TEJ. This jet is unique for the IMR. Increased cloudiness reflects more incoming solar radiation and thereby cools the region. A radiative transfer study by SBDART also confirms this mechanism. Over the AS, clouds are shallow because of the lower tropospheric thermal inversion. Hence, the LWCRF is low, but the SWCRF is highly variable, ranging from near 0 to -110 Wm-2. The high reflectivity appears to come from the large amount of aerosols transported by the low level
Study on radiative properties
65
Fig. 11 Scatter plot between the monthly mean top-of-atmosphere LWCRF (Wm-2) and SWCRF (Wm-2) from CERES/TRMM at individual grid points over the AS, BBI and EIO boxes during June to August 1998
monsoon winds from the nearby desert regions. Over the EIO, the magnitudes of the SWCRF and LWCRF are nearly the same so that they cancel each other out in an average sense. In Fig. 11, the scatter plots between the LWCRF and SWCRF over the three boxes were generated using the CERES top-of-atmosphere cloud radiative forcing data for June to August 1998. This plot also confirms that the AS and BBI boxes experience a net cooling, whereas the EIO experiences near-zero net cloud radiative forcing during the monsoon season, which matches well with the observations from the ERBS.
Appendix SBDART is a combination of a sophisticated, discrete ordinate radiative transfer module with a low-resolution atmospheric transmission model and is designed for analysis of a wide variety of RT problems encountered in satellite remote sensing. In SBDART, the radiative transfer equations are numerically integrated with a discrete ordinate radiative transfer (DISORT) module (Stamnes et al. 1988). This method uses a numerically stable algorithm to solve the equations of plane-parallel radiative transfer in a vertically inhomogeneous atmosphere (Ricchiazzi et al. 1998). In SBDART, the intensity of scattered and thermally emitted radiation can be computed at different directions and heights. Computations can be performed in up to 65 atmospheric layers and 40 radiation streams. The radiative processes included in the model are Rayleigh scattering, gaseous absorptions, cloud and aerosol scattering and absorptions. The simulation was done by keeping the cloud cover range as observed in various cloud data over the study region. The computation of radiative transfer in a cloudy atmosphere required knowledge of different
scattering parameters, which are computed in SBDART using a Mie-scattering code for spherical cloud droplets with statistical distribution of the drop radius. For radiative transfer through cloud, the parameters considered are the altitude of the cloud layer, its optical thickness and droplet size. In SBDART, altitudes of cloud layers (km) can be specified as separate cloud layers or as a range of altitudes that will be filled by cloud.
References Harrison EF, Barkstrom BR, Ramanathan V, Cess RD, Gibson G (1990) Seasonal variation of cloud radiative forcing derived from the Earth Radiation Budget Experiment. J Geophys Res 95:18.687–18.703 Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Candin L, Iredell M, Saha S, White G, Woollen J, Zhu Y, Chelliah M, Ebisuzaki W, Higgins W, Janowiak J, Mot KC, Ropelewski C, Wang Leetmaa A, Reynolds R, Jenne Roy, Joseph Dennis (1996) The NCEP/NCAR 40-Year Reanalysis Project. Bull Am Meteor Soc 77:437–471 Kiehl JT (1994) On the observed near cancellation between longwave and shortwave cloud radiative forcing in tropical regions. J Clim 7:559–565 Narayanan MS, Rao BM (1981) Detection of monsoon inversion by TIROS-N satellite. Nature 294:546–548 Rajeevan M, Srinivasan J (2000) Net cloud radiative forcing at the top of the atmosphere in the Asian monsoon region. J Clim 13:650– 657 Ricchiazzi P, Yang S, Gautier C, Sowle D (1998) SBDART: a research and teaching software tool for plane-parallel radiative transfer in the earth’s atmosphere. Bull Am Meteo Soc 79:2101– 2114 Rossow WB et al (1996) International Satellite Cloud Climatology Project (ISCCP)—Documentation of new cloud data sets. WMO/ TD-737, World Climate Research Programme (ICSU and WMO), Geneva Sathiyamoorthy V, Pal PK, Joshi PC (2004) Influence of the uppertropospheric wind shear upon cloud radiative forcing in the Asian monsoon region. J Clim 17:2725–2735
123
66 Sathiyamoorthy V, Pal PK, Joshi PC (2007) Intraseasonal variability of the tropical easterly jet. Meteorol Atmos Phy 96:305–316. doi:10.1007/s00703-006-0214-7 Stamnes K, Tsay S, Wiscombe W, Jayaweera K (1988) Numerically stable algorithm for discrete-ordinate-method radiative transfer in multiple scattering and emitting layered media. Appl Opt 27:2502–2509 Weare BC (1997) Climatic variability of cloud radiative forcing. Quart J Roy Meteor Soc 123:1055–1073 Wielicki BA, Barkstrom BR, Baum BA, Charlock TP, Green RN, Kratz DP, Lee RB III, Minnis P, Smith GL, Wong T, Young DF,
123
V. Sathiyamoorthy et al. Cess RD, Coakley JA, Crommelynck DAH, Donner L, Kandel R, King MD, Miller AJ, Ramanathan V, Randall DA, Stowe LL, Welch RM (1998) Clouds and the Earth’s radiant energy system (CERES): algorithm overview. IEEE Trans Geosci Rem Sen 36:1127–1141 Zuidema P, Mapes B (2008) Cloud vertical structure observed from space and ship over the Bay of Bengal and the Eastern Tropical Pacific. J Met Soc Jpn 86A:205–218