Search Results

You are looking at 1 - 6 of 6 items for

  • Author or Editor: Marc L. Michelsen x
  • Refine by Access: All Content x
Clear All Modify Search
Dennis L. Hartmann and Marc L. Michelsen

Abstract

Spectral analysis of a 70-year (1901–70) record of daily precipitation from 3700 stations in the country of India is carried out to search for periodicities on subseasonal time scales during the summer monsoon. Two statistically significant spectral peaks are found. A 40–50 day spectral peak corresponding to the Madden-Julian Oscillation is found over most of the portion of India south of 23°N. The phase of the oscillation is such that the precipitation maximum appears first over the relatively dry southeastern portion of the peninsula. Ten to 12 days later the precipitation peaks simultaneously all along the coast west of the Western Ghats and along a line running across India between 20° and 25°N. The precipitation maximum then spreads slowly northward and loses significance.

Cross-spectral analysis shows strong coherence between the precipitation patterns and wind oscillations. The zonal wind oscillations at 850 and 200 mb am about 180 degrees out-of-phase equatorward of about 20°N, but in-phase poleward of 20°N. Compositing of the 40–50 day variance shows the structure of the wind variations associated with the precipitation cycle over India. At the southern tip and along the west coast of India, precipitation variations on the 40–50 day time scale seem to be, at least in part, orographically controlled, with upslope winds simultaneous with the precipitation maxima. In central India the precipitation maxima are more closely related to large-scale divergence and convergence patterns, with the largest precipitation clearly associated with cyclonic circulations at low levels which have about the same horizontal structure as the precipitation maximum.

In addition to the 40–50 day oscillation, significant spectral peaks and coherent structures in precipitation are found whose characteristics correspond to those of monsoon lows. These are associated with a peak in the precipitation spectrum between about 5 and 7 days, which is strongest on the Bay of Bengal coast near 21°N. The 5–7 day oscillation in precipitation extends west-northwest across the breadth of India along the known track of monsoon lows. The phase of the oscillation indicates that the precipitation anomalies propagate westward across India with a phase speed of about 5 m s. This oscillation explains as much as 20% of the variance of precipitation in some locations, Evidence for a spectral peak near 15 days is sparse, being limited to a few regions in the north of India where the 40–50 day oscillation does not dominate the low-frequency variance.

Full access
Dennis L. Hartmann and Marc L. Michelsen

It is shown that the negative correlation between cloud-weighted sea surface temperature (SST) and high cloud fraction discussed recently by Lindzen et al. results from variations in subtropical clouds that are not physically connected to the deep convection near the equator. A negative correlation between cloud-weighted SST and average cloud fraction results from any variation in cloud fraction over the areas with lower SSTs within the domain of interest. Therefore, this correlation is not evidence that tropical cloud anvil area is inversely proportional to sea surface temperature and should not be used to infer the existence of a negative feedback in the climate system.

Full access
Dennis L. Hartmann and Marc L. Michelsen

Abstract

The dominant terms in the surface energy budget of the tropical oceans are absorption of solar radiation and evaporative cooling. If it is assumed that relative humidity in the boundary layer remains constant, evaporative cooling will increase rapidly with sea surface temperature (SST) because of the, strong temperature dependence of saturation water vapor pressure. The resulting stabilization of SST provided by evaporative cooling is sufficient to overcome positive feedback contributed by the decrease of surface net longwave cooling with increasing SST. Evaporative cooling is sensitive to small changes in boundary-layer relative humidity. Large and negative shortwave cloud forcing in the regions of highest SST are supported by the moisture convergence associated with large-scale circulations. In the descending portions of these circulations the shortwave cloud forcing is suppressed. When the effect of these circulations is taken into account by spatial averaging, the area-averaged cloud forcing shows no sensitivity to area-averaged SST changes associated with the 1987 warming event in the tropical Pacific. While the shortwave cloud forcing is large and important in the convective regions, the importance of its role in regulating the average temperature of the tropics and in modulating temperature gradients within the tropics is less clear. A heuristic model of SST is used to illustrate the possible role of large-scale atmospheric circulations on SST in the tropics and the coupling between SST gradients and mean tropical SST. The intensity of large- scale circulations responds sensitively to SST gradients and affects the mean tropical SST by supplying dry air to the planetary boundary layer. Large SST gradients generate vigorous circulations that increase evaporation and reduce the mean SST.

Full access
Dennis L. Hartmann, Marc L. Michelsen, and Stephen A. Klein

Abstract

Spectral analysis within a movable time window is used to investigate the frequency and intensity of tropical oscillations with intraseasonal time scales. Surface and rawinsonde time series from several locations in the western Pacific and Indian oceans are used. Evidence is provided that the Madden–Julian Oscillation (MJO) in the Indian Ocean region shows a seasonal variation in its preferred period from 50 days during Northern Hemisphere winter to about 35 days during summer, though this evidence is weak and stretches the limits of our ability to define a precise period for the oscillation. In the western tropical Pacific during the September–December season, a pronounced spectral peak with a central frequency between 20 and 25 days is observed. This spectral peak appears to reveal an oscillation that is distinct from the MJO but may also result from coupling between the large-scale motion and latent heat release on much smaller scales. A relationship between the oscillation and the occurrence of typhoons is demonstrated.

Full access
Dennis L. Hartmann, Maureen E. Ockert-Bell, and Marc L. Michelsen

Abstract

The role of fractional area coverage by cloud types in the energy balance of the earth is investigated through joint use of International Satellite Cloud Climatology Project (ISCCP) C1 cloud data and Earth Radiation Budget Experiment (ERBE) broadband energy flux data for the one-year period March 1985 through February 1986. Multiple linear regression is used to relate the radiation budget data to the cloud data. Comparing cloud forcing estimates obtained from the ISCCP-ERBE regression with those derived from the ERBE scene identification shows generally good agreement except over snow, in tropical convective regions, and in regions that are either nearly cloudless or always overcast. It is suggested that a substantial fraction of the disagreement in longwave cloud forcing in tropical convective regions is associated with the fact that the ERBE scene identification does not take into account variations in upper-tropospheric water vapor. On a global average basis, low clouds make the largest contribution to the net energy balance of the earth, because they cover such a large area and because their albedo effect dominates their effect on emitted thermal radiation. High, optically thick clouds can also very effectively reduce the energy balance, however, because their very high albedos overcome their low emission temperatures.

Full access
Dennis L. Hartmann, Karen J. Kowalewsky, and Marc L. Michelsen

Abstract

The scanning instruments of the Earth Radiation Budget Experiment provide measurements of instantaneous broadband albedo and outgoing longwave radiation (OLR) with a spatial resolution of about 50 km. Data from the Earth Radiation Budget Satellite (ERBS), which is in an orbit that precesses through local time at the rate of one hour every 3 days, can be used to describe the mean, hourly diurnal variations in the distribution of OLR and albedo on this scale. Much of this variation is caused by cloud type and amount changes.

Two-dimensional histograms show the co-evolution of OLR and albedo with the diurnal cycle, and the distribution of albedo–OLR pairings associated with the cloud distribution in a particular region and season. The albedo–OLR pairing characterizes a cloud type and determines its net effect on the energy balance at the top of the atmosphere. Diurnal variations in cloud type and amount in many regions are sufficient to cause substantial errors in radiation budget quantities and cloud properties estimated from observations taken from a single sun-synchronous orbit. Errors in estimated net radiation can be as lame as 50 W m−2 for oceanic stratus regions and for land regions during summer.

Full access