Search Results
You are looking at 1 - 10 of 43 items for
- Author or Editor: Bryan C. Weare x
- Refine by Access: All Content x
Abstract
Monthly means of selected variables of the 2.5° International Satellite Cloud Climatology Project (ISCCP) C2 total cloud cover (CC), cloud-top pressure (CTP), and cloud water (CW) are statistically related to sea surface temperature (SST). The statistical tools utilized include intra- and interannual correlation, regression, and composite empirical orthogonal function (EOF) analyses.
The dominant intra- and interannual composite EOFs all show that CC, CTP, and CW departures have spatially coherent links with those of SST. The second most important intra-annual functions also show coherent relations, which are about three months out of phase with those of the dominant functions. The regression analysis suggests that this phase relation may be explained by significant correlations of the cloud variables with not only SST, but also with the time derivative of SST (dSST/dt). For instance, in the tropical Pacific increased CC is accompanied by increases in SST but decreases in dSST/dt, and increased CTP is associated with decreases in SST. However, at middle and high latitudes other relationships exist, such that larger CCs may be associated with decreased SSTs, or higher CTPs may be related to higher SSTs. These diagnosed relationships have important implications for understanding cloud and cloud radiative feedbacks in weather and climate.
Abstract
Monthly means of selected variables of the 2.5° International Satellite Cloud Climatology Project (ISCCP) C2 total cloud cover (CC), cloud-top pressure (CTP), and cloud water (CW) are statistically related to sea surface temperature (SST). The statistical tools utilized include intra- and interannual correlation, regression, and composite empirical orthogonal function (EOF) analyses.
The dominant intra- and interannual composite EOFs all show that CC, CTP, and CW departures have spatially coherent links with those of SST. The second most important intra-annual functions also show coherent relations, which are about three months out of phase with those of the dominant functions. The regression analysis suggests that this phase relation may be explained by significant correlations of the cloud variables with not only SST, but also with the time derivative of SST (dSST/dt). For instance, in the tropical Pacific increased CC is accompanied by increases in SST but decreases in dSST/dt, and increased CTP is associated with decreases in SST. However, at middle and high latitudes other relationships exist, such that larger CCs may be associated with decreased SSTs, or higher CTPs may be related to higher SSTs. These diagnosed relationships have important implications for understanding cloud and cloud radiative feedbacks in weather and climate.
Abstract
Zonal averages of low, middle and total cloud amount estimates derived from measurements from Nimbus-7 have been analyzed for the six-year period April 1979 through March 1985. The globally and zonally averaged values of six-year annual means and standard deviations of total cloud amount and a proxy of cloud-top height are illustrated. Separate means for day and night and land and sea are also shown. The globally averaged value of intra-annual variability of total cloud amount is greater than 7%, and that for cloud height is greater than 0.3 km. Those of interannual variability are more than one-third of these values. Important latitudinal differences in variability are illustrated. The dominant empirical orthogonal analyses of the intra-annual variations of total cloud amount and heights show strong annual cycles, indicating that in the tropics increases in total cloud amount of up to about 30% are often accompanied by increases in cloud height of up to 1.2 km. This positive link is also evident in the dominant empirical orthogonal function of interannual variations of a total cloud/cloud height complex. This function shows a large coherent variation in total cloud cover of about 10% coupled with changes in cloud height of about 1.1 km associated with the 1982–83 El Niño–Southern Oscillation event.
Abstract
Zonal averages of low, middle and total cloud amount estimates derived from measurements from Nimbus-7 have been analyzed for the six-year period April 1979 through March 1985. The globally and zonally averaged values of six-year annual means and standard deviations of total cloud amount and a proxy of cloud-top height are illustrated. Separate means for day and night and land and sea are also shown. The globally averaged value of intra-annual variability of total cloud amount is greater than 7%, and that for cloud height is greater than 0.3 km. Those of interannual variability are more than one-third of these values. Important latitudinal differences in variability are illustrated. The dominant empirical orthogonal analyses of the intra-annual variations of total cloud amount and heights show strong annual cycles, indicating that in the tropics increases in total cloud amount of up to about 30% are often accompanied by increases in cloud height of up to 1.2 km. This positive link is also evident in the dominant empirical orthogonal function of interannual variations of a total cloud/cloud height complex. This function shows a large coherent variation in total cloud cover of about 10% coupled with changes in cloud height of about 1.1 km associated with the 1982–83 El Niño–Southern Oscillation event.
Abstract
Regional estimates of low, middle, high, and total cloud amounts derived from bispectral measurements from Nimbus-7 have been analyzed for the six-year period April 1979 through March 1985. Fractional cloud cover for the three height categories was used to calculate a proxy mean cloud-top height. Intra- and interannual standard deviations of total cloud amount and mean cloud height show realistic patterns throughout most of the globe except at very high latitudes. Over much of the cash, intra-annual and interannual variations in total cloud amount are strongly positively correlated with variations in cloud height. Furthermore, both total cloud amount and cloud height variations are moderately correlated with sea surface temperature variations. The strongest correlations are positive in the tropics for both intra-annual and interannual variations. In middle latitudes, moderate negative correlations are positive with intra-annual variations, whereas moderate positive correlations occur on interannual time frames. In the tropics 1°C changes in temperature are statistically related to a change of total cloudiness of at least 2% and a change in cloud height of more than 0.5 km.
Abstract
Regional estimates of low, middle, high, and total cloud amounts derived from bispectral measurements from Nimbus-7 have been analyzed for the six-year period April 1979 through March 1985. Fractional cloud cover for the three height categories was used to calculate a proxy mean cloud-top height. Intra- and interannual standard deviations of total cloud amount and mean cloud height show realistic patterns throughout most of the globe except at very high latitudes. Over much of the cash, intra-annual and interannual variations in total cloud amount are strongly positively correlated with variations in cloud height. Furthermore, both total cloud amount and cloud height variations are moderately correlated with sea surface temperature variations. The strongest correlations are positive in the tropics for both intra-annual and interannual variations. In middle latitudes, moderate negative correlations are positive with intra-annual variations, whereas moderate positive correlations occur on interannual time frames. In the tropics 1°C changes in temperature are statistically related to a change of total cloudiness of at least 2% and a change in cloud height of more than 0.5 km.
Abstract
Multilag singular value decomposition (MLSVD) analysis is developed and applied to diagnosing the impact of interannual variations of outgoing longwave radiation (OLR) on tropical stratospheric temperature changes. MLSVD is designed to analyze simultaneously variations at multiple levels and for a large number of temporal lags and leads. The two dominant MLSVDs are strongly related to El Niño–Southern Oscillation (ENSO). The associated patterns of tropical OLR are similar to the canonical ENSO SST patterns with strong negative sign regions stretching along the equator in the eastern and central Pacific. These dominant modes are strongly linked to temperature perturbations at a wide range of lags. At the lowest analyzed level (200 hPa) and zero lag positive temperatures anomalies are in the region of low OLR. In the lower stratosphere near 100 hPa, strong negative temperature perturbations replace the positive values of the lowest level. Higher in the stratosphere near 20 hPa, equatorial temperature perturbations are again positive, but with a more zonally elongated spatial pattern. Overall, the equatorial temperature anomalies propagate slowly to the east, at a speed strongly related to ocean–atmosphere coupling of less than 1 m s−1, and vertically and westward into the stratosphere by Rossby waves with a speed in the range of 30 m s−1.
Abstract
Multilag singular value decomposition (MLSVD) analysis is developed and applied to diagnosing the impact of interannual variations of outgoing longwave radiation (OLR) on tropical stratospheric temperature changes. MLSVD is designed to analyze simultaneously variations at multiple levels and for a large number of temporal lags and leads. The two dominant MLSVDs are strongly related to El Niño–Southern Oscillation (ENSO). The associated patterns of tropical OLR are similar to the canonical ENSO SST patterns with strong negative sign regions stretching along the equator in the eastern and central Pacific. These dominant modes are strongly linked to temperature perturbations at a wide range of lags. At the lowest analyzed level (200 hPa) and zero lag positive temperatures anomalies are in the region of low OLR. In the lower stratosphere near 100 hPa, strong negative temperature perturbations replace the positive values of the lowest level. Higher in the stratosphere near 20 hPa, equatorial temperature perturbations are again positive, but with a more zonally elongated spatial pattern. Overall, the equatorial temperature anomalies propagate slowly to the east, at a speed strongly related to ocean–atmosphere coupling of less than 1 m s−1, and vertically and westward into the stratosphere by Rossby waves with a speed in the range of 30 m s−1.
Abstract
Diabatic heating rates calculated in the UCLA general circulation model are analyzed. The heating rates are for the December–January–February season for the mean of four climatological runs and the mean of four runs with observed 1982/83 sea surface temperatures as boundary conditions. Vertically integrated total diabatic heating for the climatological calculations agrees well with observations. Unfortunately, the available observations of zonally averaged vertical structure differ from each other so substantially that comparisons with the model are inconclusive. Nevertheless, the vertical structure of the model seems quite realistic and in general agreement with analyses of tropical cloud clusters or middle latitude cyclones.
The model diabatic heating is stratified in a number of ways. Zonal average height-latitude cross sections of the heating due to cumulus and large-scale precipitation and radiation show complex profiles for the rates associated with precipitation and much less structure with those associated with radiation except near the ground. Sample vertical profiles for regions in the tropics and new 45°N suggest that the zonal averages are generally representative of the individual heating profiles except near the surface.
Anomalous heating rates associated with the model 1982/83 season, in which a a strong El Niño was observed, show a moderate alteration in diabatic heating over the breadth of the tropical Pacific. In the eastern equatorial Pacific, increases are evident for the heating rates associated with cumulus and large-scale precipitation and radiation. Reductions in cumulus and radiative heating over other parts of the tropics are also apparent.
Abstract
Diabatic heating rates calculated in the UCLA general circulation model are analyzed. The heating rates are for the December–January–February season for the mean of four climatological runs and the mean of four runs with observed 1982/83 sea surface temperatures as boundary conditions. Vertically integrated total diabatic heating for the climatological calculations agrees well with observations. Unfortunately, the available observations of zonally averaged vertical structure differ from each other so substantially that comparisons with the model are inconclusive. Nevertheless, the vertical structure of the model seems quite realistic and in general agreement with analyses of tropical cloud clusters or middle latitude cyclones.
The model diabatic heating is stratified in a number of ways. Zonal average height-latitude cross sections of the heating due to cumulus and large-scale precipitation and radiation show complex profiles for the rates associated with precipitation and much less structure with those associated with radiation except near the ground. Sample vertical profiles for regions in the tropics and new 45°N suggest that the zonal averages are generally representative of the individual heating profiles except near the surface.
Anomalous heating rates associated with the model 1982/83 season, in which a a strong El Niño was observed, show a moderate alteration in diabatic heating over the breadth of the tropical Pacific. In the eastern equatorial Pacific, increases are evident for the heating rates associated with cumulus and large-scale precipitation and radiation. Reductions in cumulus and radiative heating over other parts of the tropics are also apparent.
Abstract
Zonal average temperature for each 15° latitude band of the globe for six atmospheric levels is computed for crush mouth between May 1958 and December 1974. The averages are based on reports at a total of 190 WMO stations. Monthly and annual means, interannual variances and one-month lag correlations are presented. The correlations between zonal mean temperatures at different levels within the same latitudinal zone and different zones at the same level are also illustrated. Sample power spectra are presented for middle-latitude and subtropical zonal mean temperatures.
Abstract
Zonal average temperature for each 15° latitude band of the globe for six atmospheric levels is computed for crush mouth between May 1958 and December 1974. The averages are based on reports at a total of 190 WMO stations. Monthly and annual means, interannual variances and one-month lag correlations are presented. The correlations between zonal mean temperatures at different levels within the same latitudinal zone and different zones at the same level are also illustrated. Sample power spectra are presented for middle-latitude and subtropical zonal mean temperatures.
Abstract
Interannual variations of net surface heating of the tropical Pacific Ocean are analyzed for the period 1957–76. Special emphasis is given to exploring the relationship between these variations and those in sea temperature associated with El Niño/Southern Oscillation. The analyses include eigenvector analysis, composites of the net heating for various phase of El Niño, and time series analysis of various measures of the variability. The results indicate large-scale patterns of variability dominated by time scales greater than one year. A link between the large-scale variations of sea temperature and net surface heating is evident from each set of analyses. In general, anomalously high heating appears to be associated with cold water. However, it also seems apparent that greater than average heating of the ocean persists for several months into the pe6ods of El Niño when sea temperatures are rising rapidly. Thus it is hypothesized that net surface heating contributes to the development of the early stages of an El Niño warm period.
Abstract
Interannual variations of net surface heating of the tropical Pacific Ocean are analyzed for the period 1957–76. Special emphasis is given to exploring the relationship between these variations and those in sea temperature associated with El Niño/Southern Oscillation. The analyses include eigenvector analysis, composites of the net heating for various phase of El Niño, and time series analysis of various measures of the variability. The results indicate large-scale patterns of variability dominated by time scales greater than one year. A link between the large-scale variations of sea temperature and net surface heating is evident from each set of analyses. In general, anomalously high heating appears to be associated with cold water. However, it also seems apparent that greater than average heating of the ocean persists for several months into the pe6ods of El Niño when sea temperatures are rising rapidly. Thus it is hypothesized that net surface heating contributes to the development of the early stages of an El Niño warm period.
Abstract
The spatial and temporal character of El Niño is explored with analyses of tropical Pacific Ocean surface temperatures for the period 1957–76. The data are derived from approximately 5×106 marine weather reports. Maps are illustrated which portray the initiation, maturation and decay of an “average” El Niño event. Empirical orthogonal functions of nonseasonal departures are displayed. The time coefficients of the dominant empirical functions are derived together with average departures for 18 regions which are usually 10° of latitude and 40–50° of longitude in size. Lag correlation and coherence-spectral analysts are carried out on all of the time series. The pattern of El Niño which is portrayed is that of a basinwide phenomenon with a time evolution lasting more than 24 months. During this evolution sea temperatures in the western Pacific tend to have departures of opposite sign to those in the cast. Variations in the eastern equatorial region are shown to precede those in the central equatorial and northeastern tropical Pacific by 1–4 months. On the other hand, changes in the central and eastern Pacific near 25°S tend to precede those in the equatorial region by a few months.
Abstract
The spatial and temporal character of El Niño is explored with analyses of tropical Pacific Ocean surface temperatures for the period 1957–76. The data are derived from approximately 5×106 marine weather reports. Maps are illustrated which portray the initiation, maturation and decay of an “average” El Niño event. Empirical orthogonal functions of nonseasonal departures are displayed. The time coefficients of the dominant empirical functions are derived together with average departures for 18 regions which are usually 10° of latitude and 40–50° of longitude in size. Lag correlation and coherence-spectral analysts are carried out on all of the time series. The pattern of El Niño which is portrayed is that of a basinwide phenomenon with a time evolution lasting more than 24 months. During this evolution sea temperatures in the western Pacific tend to have departures of opposite sign to those in the cast. Variations in the eastern equatorial region are shown to precede those in the central equatorial and northeastern tropical Pacific by 1–4 months. On the other hand, changes in the central and eastern Pacific near 25°S tend to precede those in the equatorial region by a few months.
Abstract
A shallow water equations model of the tropical atmosphere is developed and tested. The model includes a simple parameterization of convective rainfall so that the model heating is a function of fixed sea surface temperatures and model dependent circulation. To achieve quasi-steady states the model is iterated a number of steps.
For equatorial sea temperature perturbations imposed on an idealized Pacific Ocean, the model shows two distinct modes of response. In one mode local evaporation is slightly more important than moisture convergence. This results in a relatively weak, equatorially trapped perturbation circulation. In the other mode moisture convergence is dominant and the associated circulation is stronger and less equatorially trapped. The necessary condition for the second “enhanced” mode is that the positive sea temperature perturbations overlap sufficiently with the climatologically warm waters of the western ocean.
Subtropical sea temperature perturbations generally result in a steady state heating with a relatively broad meridional scale and moderate magnitude. The final states are relatively insensitive to the position of the sea temperature perturbations relative to the mean sea temperature field.
While these results must be viewed as quite preliminary due to the simple nature of the dynamical model, they do suggest that actual atmospheric heating perturbations may be the result of complex interactions between oceanic and atmospheric conditions by way of both the surface evaporation and moisture convergence processes.
Abstract
A shallow water equations model of the tropical atmosphere is developed and tested. The model includes a simple parameterization of convective rainfall so that the model heating is a function of fixed sea surface temperatures and model dependent circulation. To achieve quasi-steady states the model is iterated a number of steps.
For equatorial sea temperature perturbations imposed on an idealized Pacific Ocean, the model shows two distinct modes of response. In one mode local evaporation is slightly more important than moisture convergence. This results in a relatively weak, equatorially trapped perturbation circulation. In the other mode moisture convergence is dominant and the associated circulation is stronger and less equatorially trapped. The necessary condition for the second “enhanced” mode is that the positive sea temperature perturbations overlap sufficiently with the climatologically warm waters of the western ocean.
Subtropical sea temperature perturbations generally result in a steady state heating with a relatively broad meridional scale and moderate magnitude. The final states are relatively insensitive to the position of the sea temperature perturbations relative to the mean sea temperature field.
While these results must be viewed as quite preliminary due to the simple nature of the dynamical model, they do suggest that actual atmospheric heating perturbations may be the result of complex interactions between oceanic and atmospheric conditions by way of both the surface evaporation and moisture convergence processes.
Abstract
The role of moisture variations in the initiation of Madden–Julian oscillation (MJO) variability is examined. Composite singular value decomposition (CSVD) analyses at various time lags are carried out to discern the complex space–time relationships between convection, low-level specific humidity and divergence, and surface evaporation. The utilized data are low-level moisture and winds and surface evaporation from the NCEP–NCAR reanalysis and outgoing longwave radiation (OLR) from NOAA satellite observations for the period 1981–2000. These data are filtered at each point using a 150-point Lanczos filter capturing well the 20–100-day periodicities.
The two dominant CSVDs describe an eastward-propagating quasi-wavenumber-1 system in all of the analyzed variables. The dominant low-level divergence, moisture, and latent heat flux variations lead those of OLR by approximately 10 days. The low-level convergence and positive moisture anomalies develop near and to the east of the ensuing convective perturbations; positive latent heat flux variations generally occur farther to the east. Moisture variations develop at least 15 days prior to a convective event and have stronger correlations with OLR than those of simultaneous low-level divergence. Near the centers of convection the low-level moisture increases at the same time that the 1000-hPa flow is becoming slightly more divergent. This implies that the moisture preconditioning of convective events is not driven totally by moisture convergence. This may require a modification of the frictional wave conditional instability of the second kind (CISK) hypothesis for MJO development.
Abstract
The role of moisture variations in the initiation of Madden–Julian oscillation (MJO) variability is examined. Composite singular value decomposition (CSVD) analyses at various time lags are carried out to discern the complex space–time relationships between convection, low-level specific humidity and divergence, and surface evaporation. The utilized data are low-level moisture and winds and surface evaporation from the NCEP–NCAR reanalysis and outgoing longwave radiation (OLR) from NOAA satellite observations for the period 1981–2000. These data are filtered at each point using a 150-point Lanczos filter capturing well the 20–100-day periodicities.
The two dominant CSVDs describe an eastward-propagating quasi-wavenumber-1 system in all of the analyzed variables. The dominant low-level divergence, moisture, and latent heat flux variations lead those of OLR by approximately 10 days. The low-level convergence and positive moisture anomalies develop near and to the east of the ensuing convective perturbations; positive latent heat flux variations generally occur farther to the east. Moisture variations develop at least 15 days prior to a convective event and have stronger correlations with OLR than those of simultaneous low-level divergence. Near the centers of convection the low-level moisture increases at the same time that the 1000-hPa flow is becoming slightly more divergent. This implies that the moisture preconditioning of convective events is not driven totally by moisture convergence. This may require a modification of the frictional wave conditional instability of the second kind (CISK) hypothesis for MJO development.
