Wind-driven latent heat flux and the intraseasonal oscillation

Transcription

1 GEOPHYSICAL RESEARCH LETTERS, VOL. 35, L04815, doi: /2007gl032746, 2008 Wind-driven latent heat flux and the intraseasonal oscillation Nilesh M. Araligidad 1 and Eric D. Maloney 1,2 Received 20 November 2007; revised 14 January 2008; accepted 28 January 2008; published 26 February [1] The importance of tropical west Pacific wind-driven latent heat flux anomalies for supporting boreal winter intraseasonal precipitation variability is analyzed during using satellite and in-situ observations. Intraseasonal ( day) wind speed anomalies from QuikSCAT are significantly correlated with TRMM precipitation anomalies, with instantaneous correlations peaking near 0.7 in the regions of strongest west Pacific intraseasonal precipitation variance. Positive intraseasonal wind speed anomalies occur within regions of enhanced precipitation during intraseasonal oscillation (ISO) events, suggesting an increase in the wind-driven component of latent heat flux then. Consistent with these results, west Pacific intraseasonal TAO buoy latent heat flux and TRMM precipitation anomalies are significantly correlated ( ), and latent heat flux anomalies are primarily winddriven. Collocated evaporation anomalies are approximately 20% of intraseasonal precipitation anomalies. Intraseasonal precipitation in the west Pacific may be supported by winddriven surface fluxes, consistent with the modeling work of Maloney and Sobel. Citation: Araligidad, N. M., and E. D. Maloney (2008), Wind-driven latent heat flux and the intraseasonal oscillation, Geophys. Res. Lett., 35, L04815, doi: /2007gl Introduction [2] The intraseasonal oscillation (ISO) is the dominant mode of tropical intraseasonal variability. The ISO is manifest as an eastward propagating, zonal wavenumber 1 3 disturbance in winds and precipitation with characteristic timescales of days [e.g., Madden and Julian, 2005]. The greatest amplitude ISO precipitation variations occur in the east Indian and west Pacific Oceans during boreal winter, where coupling to the large-scale circulation is strong. Early theoretical studies hypothesized that winddriven surface fluxes might be important to the ISO [e.g., Emanuel, 1987; Neelin et al., 1987]. Such early models assumed mean easterly boundary layer flow where lowlevel easterly anomalies enhance latent heat flux to the east of ISO convection, increasing boundary layer moist static energy and supporting eastward propagation of ISO convection. These early studies were generally criticized on the grounds that enhanced evaporation in the observed ISO generally occurs where the surface wind anomalies have a westerly component [e.g., Jones and Weare, 1996; Lau and Sui, 1997; Zhang and McPhaden, 2000]. 1 College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, Oregon, USA. 2 Now at Department of Atmospheric Science, Colorado State University, Fort Collins, Colorado, USA. Copyright 2008 by the American Geophysical Union /08/2007GL [3] The lack of observational support for early windevaporation feedback models of the ISO does not rule out that surface fluxes are important for ISO maintenance. Recent modeling evidence suggests that wind-induced surface fluxes are important for producing realistic tropical intraseasonal variability, and many recent models have been able to simulate the observed relationship between precipitation and anomalous latent heat flux [e.g., Lin et al., 2000; Raymond, 2001; Maloney and Sobel, 2004]. Maloney and Sobel [2004] found that west Pacific ISO convection in their model decreased in amplitude by nearly 50% when the wind-induced latent heat flux anomalies were removed by fixing surface latent heat fluxes to climatology. The present study develops observational evidence support for the hypothesis that wind-induced surface heat fluxes are important for supporting ISO convection. [4] Our study further suggests the importance of windinduced surface heat fluxes to the column integrated moist static energy budget [e.g., Neelin and Held, 1987]. Assuming a deep convective heating profile, Yu et al. [1998] showed that vertical circulations export moist static energy at rate of approximately 16 20% of precipitation over the west Pacific warm pool. Observational estimates suggest that anomalous column-integrated cloud-radiative heating is about 10 15% of intraseasonal precipitation in the west Pacific [e.g., Lin and Mapes, 2004], suggesting that cloudradiative feedbacks alone are not enough to support a radiative-convective instability. In this study, we derive similar quantitative estimates of the relationship of intraseasonal latent heat flux to precipitation using west Pacific Tropical Ocean Global Atmosphere (TOGA) Tropical Atmosphere Ocean (TAO) buoys and Tropical Rainfall Measuring Mission (TRMM) precipitation, with the acknowledgement that the deep convective structure assumptions of Yu et al. [1998] are an over-idealization since diabatic heating profiles vary substantially over an ISO lifecycle [e.g., Kiladis et al., 2005]. [5] Our study examines the relationship between intraseasonal TRMM precipitation and TAO buoy latent heat fluxes at 5 S, 165 E and 8 S, 165 E during These locations are coincident with the strongest west Pacific ISO convective variability during austral summer, and have a sufficiently long record to allow statistical significance to be determined. Most studies that have examined the relationship of surface evaporation to ISO convection have relied on model analyses to determine fluxes [e.g., Jones and Weare, 1996; Shinoda et al., 1998; Sperber, 2003], or on in-situ measurements with data records less than a year [e.g., Lau and Sui, 1997]. Zhang [1996] and Zhang and McPhaden [2000] examined the evolution of the surface heat budget using TAO buoy data during ISO events, although Zhang [1996] examined fluxes during a limited three year period ( ), and Zhang and McPhaden [2000] concentrated on buoy fluxes L of5

2 within two degrees of the equator. We also use the Quick Scatterometer (QuikSCAT) ocean vector wind product during to determine how wind speed variations are related to ISO precipitation. 2. Data and Methods 2.1. TAO Buoy data [6] Atmospheric and oceanic fields from TAO buoys at 5 S, 165 E and 8 S, 165 E [McPhaden, 1995] are used to compute latent heat flux. Using daily averaged fields of relative humidity, air temperature, wind speed and one meter depth sea surface temperature during , surface latent heat fluxes are calculated using version 3.0 of the Coupled Ocean-Atmosphere Response Experiment (COARE) flux bulk algorithm [Fairall et al., 2003] QuikSCAT Ocean Vector Winds [7] The SeaWinds Scatterometer on the NASA QuikSCAT satellite retrieves ocean surface wind stress from the ocean surface using the backscatter from multiple azimuth and incidence angles [Chelton and Freilich, 2005]. QuikSCAT vector winds were downloaded on a grid in daily ascending and descending swaths from Remote Sensing Systems of Santa Rosa, California ( for the period Because rain contamination of the backscatter signal can occur, QuikSCAT data are considered missing and removed when Special Sensor Microwave Imager radiometer data [e.g., Wentz and Spencer, 1998] indicates that it is raining within, or adjacent to, the scatterometer grid cell. Except where noted in the auxiliary material, 1 wind speed is calculated in-swath using vector wind components on the grid. Daily means of all wind-related fields were then constructed from ascending and descending swaths, and then averaged to a 1 1 grid before further processing. Wind speed and vector winds components are then composited as overlapping 3-day averages at daily intervals to further mitigate the effects of missing data from the sampling pattern of the QuikSCAT satellite and from rain contamination of individual measurements. Linear interpolation in time then fills remaining data gaps. Missing data maximize at 0.3% of the total record in regions of high mean precipitation before interpolation TRMM Precipitation Data [8] Daily-averaged precipitation fields during from the TRMM Level 3B-42 Version 6 product were downloaded from the Goddard Space Flight Center Distributed Access Archive System (accessible from trmm.gsfc.nasa.gov/), and then averaged to a 1 1 grid [e.g., Huffman et al., 2001]. TRMM data are interpolated to individual buoy locations for comparison with TAO intraseasonal latent heat fluxes. 3. Results 3.1. ISO Precipitation and Wind Anomalies [9] To examine where strong precipitation anomalies occur during ISO events, and to examine the general relationship between ISO surface wind and precipitation 1 Auxiliary materials are available in the HTML. doi: / 2007GL (P) anomalies in the west Pacific during austral summer, we generate a composite ISO event using the index and composting method of Maloney and Kiehl [2002]. This index is derived from the first two empirical orthogonal functions of equatorial intraseasonal ( day) 850 hpa zonal wind (hu 850 i) from the National Centers for Environmental Prediction/National Center for Atmospheric Research reanalysis I [Kalnay et al., 1996]. Hereafter, a field f bandpass filtered to days will be denoted by h f i. Two sixty-point linear non-recursive filters with halfpower points at 30 and 100 days are used to construct this intraseasonal bandpass filter. Sixteen ISO events were isolated during November April of using an event selection threshold of one standard deviation from zero, and data were averaged across all events to create a composite. [10] Composite anomalous precipitation and wind vector are displayed for the enhanced precipitation phase of the ISO over the west Pacific warm pool (Figure 1, top), where ~u sfc =(u sfc, v sfc ) is the surface vector wind. During the enhanced precipitation phase, positive bandpassed precipitation hpi peaks between 0 and 10 S, and is associated with strong westerly hu sfc i near and to the west of positive hpi. The region of strong hpi during ISO events is consistent with locations where overall hpi variance is enhanced during November April (Figure S1 of the auxiliary material). An exception is in the Intertropical Convergence Zone of the North Pacific where composite hpi is of low amplitude. [11] Anomalous wind speed hspdi and h~u sfc i during the enhanced precipitation phase of the ISO are shown in Figure 1 (bottom). Positive hspdi occurs near and to the west of ISO convection. The behavior shown in Figure 1 is similar for periods of suppressed precipitation, although with opposite-signed anomalies (not shown). The relationship between composite hspdi and hpi shown in Figure 1 suggests that wind-driven latent heat flux anomalies may support ISO convection in these regions, as verified by the buoy analysis below. [12] Text S1 of the auxiliary material describes the factors that regulate hspdi. The enhancement (suppression) of wind speed during periods of enhanced (suppressed) ISO precipitation appears to be primarily due to the addition of hu sfc i and hv sfc i to the mean flow. Low-level westerly (easterly) hu sfc i superimposed on the background near-equatorial Southern Hemisphere flow that is mean westerly to the west of 170 E is primarily responsible for the positive (negative) hspdi there. However, northerly (southerly) hv sfc i to the east of 170 E is primarily responsible for the eastward extension of positive (negative) hspdi past 170 E. hv sfc i is also important to hspdi at other stages of an ISO lifecycle. Intraseasonal variations in high frequency transient activity such as easterly waves and mesoscale gustiness make only modest contributions to the wind speed anomalies, with the largest contributions occurring on the poleward flanks of the vector wind anomaly maxima in Figure 1 (bottom) in regions of anomalous vorticity Correlation of Day Wind-Speed and Precipitation Anomalies [13] The correlation of hspdi and hpi suggests whether the wind-induced component of latent heat flux supports intraseasonal convection. hspdi and hpi show a significant 2of5

3 Figure 1. Composite (top) hpi and h~u sfc i, and (bottom) hspdi and h~u sfc i for the enhanced precipitation phase of the ISO. The reference wind vector is shown at the bottom right (m s 1 ). Precipitation units are mm day 1. Wind speed units are m s 1. The two buoy locations used in Figure 3 are indicated by the blue markers in Figure 1 (top). positive instantaneous correlation in regions where strong hpi occurs during ISO events (Figure 2). The highest correlations that peak near 0.7 generally occur to the west of 170 E where the mean flow (vectors in Figure 2) is generally weak westerly. The t-statistic demonstrates that correlation coefficients greater than 0.3 are significantly different from zero at 90% confidence level. Since the mean period of the ISO is about 40 days, we conservatively use N/40 as the number of degrees of freedom, where N represents the number of daily observations at a particular location. [14] The correlations of Figure 2 were also recalculated by first averaging vector wind components to a grid before computing a wind speed as the magnitude of this vector. Differences in the hpi vs. hspdi correlation using this method as compared to Figure 2 would suggest whether mesoscale gustiness and other high wavenumber spatial variability contributes to the positive correlation of hspdi and hpi [e.g., Back and Bretherton, 2005]. Only modest decreases to the correlation coefficients in Figure 2 occur (0.1) when wind speed is calculated on the grid, suggesting that mesoscale gustiness is of second order importance in regulating the hpi vs. hspdi correlations Correlation of Intraseasonal Latent Heat Flux and Precipitation Anomalies [15] As discussed above, wind speed fields determined from QuikSCAT can only be used to determine where hspdi would affect the wind-driven part of the latent heat flux, and give no information about the actual latent heat flux. TAO buoys provide meteorological and oceanographic measurements that can be used to compute surface heat fluxes in total. [16] We generate scatter plots of intraseasonal TRMM precipitation anomalies versus TAO latent heat flux anomalies at the 5 S, 165 E and 8 S, 165 E TAO buoys during November April (Figure 3). These buoy locations are in the band of strongest ISO precipitation variability in the west Pacific during November April (e.g. see buoy locations on Figure 1, top). Daily latent heat flux can be calculated for 1163 days at 5 S, 165 E and 1099 days at 8 S, 165 E during November April of Due to the intermittent data record at the TAO buoys, to calculate intraseasonal anomalies of latent heat flux and precipitation for the scatterplots we use a bandpass filter modified from that used in Figures 1 and 2. The modified filter is described in the auxiliary material. [17] Correlation coefficients between intraseasonal precipitation and latent heat flux are 0.47 at 5 S, 165 E and 0.59 at 8 S, 165 E. These correlation coefficients are statistically significant at the 90% confidence level. These correlations increase by 0.1 to 0.2 if intraseasonal air-sea humidity difference anomalies are removed before calculation of latent heat flux, thus making correlation magnitudes consistent with the wind speed-precipitation correlations of Figure 2. The regression coefficients between intraseasonal precipitation and latent heat flux are 4.1 (W m 2 3of5

4 Figure 2. Instantaneous correlation of hspdi and hpi during November April. Mean November April QuikSCAT wind vectors are also shown. The reference wind vector in m s 1 is located at the bottom right. Stippling indicates where the correlation coefficient is statistically significant at the 90% confidence level. (W m 2 ) 1 )at5 S, 165 E and 5.5 at 8 S, 165 E TAO buoys. These coefficients indicate that intraseasonal latent heat flux anomalies are about 18 24% of precipitation anomalies in the west Pacific warm pool in regions where ISO precipitation variance maximizes during boreal winter. Recall that Yu et al. [1998] estimated that deep convective circulations export moist static energy at rate of approximately 16 20% of precipitation anomalies over the west Pacific warm pool. Thus, it appears that latent heat flux anomalies are comparably important to the moist static energy budget of the ISO. [18] Sensitivity tests involving linearization of the bulk formula for latent heat flux in which SST and near-surface humidity are fixed at their 60-day running average indicate that November April intraseasonal latent heat flux anomalies in the west Pacific are primarily wind-driven. In fact, thermodynamic contributions to the latent heat flux anomaly reduce the amplitude by about 20% (not shown here), consistent with previous studies that show intraseasonal SST anomalies reduce the amplitude of surface heat fluxes during ISO events [e.g., Shinoda et al., 1998]. 4. Conclusions [19] The relationship between intraseasonal latent heat flux and precipitation is examined during November April of using satellite and buoy observations in the tropical west Pacific. QuikSCAT surface winds are used in tandem with TRMM precipitation to show that intraseasonal oscillation (ISO) enhanced precipitation phases are associated with positive day wind speed anomalies, and suppressed precipitation phases are associated with negative wind speed anomalies. Zonal and meridional intraseasonal vector wind anomalies interacting with the mean flow are primarily responsible for this modulation of intraseasonal wind speed. High frequency (less than 20 day timescale) components of the wind field and their intraseasonal variability are less important to intraseasonal wind speed anomalies, and their influence maximizes on the poleward flanks of the strongest intraseasonal precipitation and vector wind anomalies. [20] Intraseasonal QuikSCAT wind speed and TRMM precipitation anomalies exhibit statistically significant Figure 3. Scatterplot of intraseasonal TRMM precipitation vs. TAO buoy latent heat flux at (top) 5 S, 165 E and (b) 8 S, 165 E. Plot details are described in the text and auxiliary material. 4of5

5 correlations (>0.7) in regions of strong ISO precipitation variability, suggesting that intraseasonal precipitation anomalies are associated with an increase in the winddriven component of latent heat flux. [21] TRMM precipitation and TAO buoy latent heat fluxes at 5 S, 165 E and 8 S, 165 E are significantly correlated; with instantaneous correlations of Sensitivity tests indicate that these latent heat flux anomalies are dominated by the wind-driven component. Intraseasonal latent heat flux anomalies are about 20% of precipitation, a significant perturbation to the column-integrated moist static energy budget of the ISO. The results of this observational study lend support to the hypothesis of Maloney and Sobel [2004] that wind-induced surface heat fluxes in the west Pacific warm pool are an important regulator of convection during ISO events. Future work will necessarily examine how intraseasonal latent heat flux anomalies contribute to a positive correlation of diabatic heating and tropospheric temperature anomalies, generating eddy available potential energy that may help support the large-scale ISO circulation against dissipation. [22] Acknowledgments. This research supported under grants ATM , ATM , and ATM from the Climate and Large- Scale Dynamics Program of the National Science Foundation. The statement, findings, conclusions, and recommendations do not necessarily reflect the views of NSF. References Back, L. E., and C. S. Bretherton (2005), The relationship between wind speed and precipitation in the east Pacific ITCZ, J. Clim., 18, Chelton, D. B., and M. H. Freilich (2005), Scatterometer-based assessment of 10-m wind analyses from the operational ECMWF and NCEP numerical weather prediction models, Mon. Weather Rev., 133, Emanuel, K. A. (1987), An air-sea interaction model of intraseasonal oscillations in the tropics, J. Atmos. Sci., 44, Fairall, C. W., E. F. Bradley, J. E. Hare, A. A. Grachev, and J. B. Edson (2003), Bulk parameterization of air-sea fluxes: Updates and verification for the COARE algorithm, J. Clim., 16, Huffman, G. H., R. F. Adler, M. M. Morrissey, D. T. Bolvin, S. Curtis, R. Joyce, B. McGavock, and J. Susskind (2001), Global precipitation at one-degree daily resolution from multisatellite observations, J. Hydrometeorol., 2, Jones, C., and B. C. Weare (1996), The role of low-level moisture convergence and ocean latent heat fluxes in the Madden-Julian oscillation: An observational analysis using ISCCP data and ECMWF analyses, J. Clim., 11, Kalnay, E., et al. (1996), The NCEP/NCAR 40-year reanalysis project, Bull. Am. Meteorol. Soc., 77, Kiladis, G. N., K. H. Straub, and P. T. Haertel (2005), Zonal and vertical structure of the Madden Julian oscillation, J. Atmos. Sci., 62, Lau, K. M., and C. H. Sui (1997), Mechanism of short-term sea surface temperature regulation: Observations during TOGA COARE, J. Clim., 10, Lin, J., and B. E. Mapes (2004), Radiation budget of the tropical intraseasonal oscillation, J. Atmos. Sci., 61, Lin, J. W.-B., J. D. Neelin, and N. Zeng (2000), Maintenance of tropical intraseasonal variability: Impact of evaporation-wind feedback and midlatitude storms, J. Atmos. Sci., 57, Madden, R., and P. R. Julian (2005), Historical perspective, in Intraseasonal Variability in the Atmosphere-Ocean Climate System, edited by K. M. Lau and D. E. Waliser, chap. 1, pp. 1 18, Springer, Berlin. Maloney, E. D., and J. T. Kiehl (2002), MJO-related SST variations over the tropical Eastern Pacific during Northern Hemisphere summer, J. Clim., 15, Maloney, E. D., and A. H. Sobel (2004), Surface fluxes and ocean coupling in the tropical intraseasonal oscillation, J. Clim., 17, McPhaden, M. J. (1995), The Tropical Atmosphere Ocean array is completed, Bull. Am. Meteorol. Soc., 76, Neelin, J. D., and I. M. Held (1987), Modeling tropical convergence based on the moist static energy budget, Mon. Weather Rev., 115, Neelin, J. D., I. M. Held, and K. H. Cook (1987), Evaporation-wind feedback and low frequency variability in the tropical atmosphere, J. Atmos. Sci., 44, Raymond, D. J. (2001), A new model of the Madden-Julian Oscillation, J. Atmos. Sci., 58, Shinoda, T., H. H. Hendon, and J. Glick (1998), Intraseasonal variability of surface fluxes and sea surface temperature in the tropical Indian and Pacific oceans, J. Atmos. Sci., 11, Sperber, K. R. (2003), Propagation and vertical structure of the Madden- Julian Oscillations, Mon. Weather Rev., 131, Wentz, F. J., and R. W. Spencer (1998), SSM/I rain retrievals within a unified all-weather ocean algorithm, J. Atmos. Sci., 55, Yu, J.-Y., C. Chou, and J. D. Neelin (1998), Estimating the gross moist stability of the tropical atmosphere, J. Atmos. Sci., 55, Zhang, C. (1996), Atmospheric intraseasonal variability at the surface in the western Pacific Ocean, J. Atmos. Sci., 53, Zhang, G. J., and M. J. McPhaden (2000), Intraseasonal surface cooling in the equatorial western Pacific, J. Clim., 13, N. M. Araligidad, College of Oceanic and Atmospheric Sciences, Oregon State University, 104 COAS Administrative Building, Corvallis, OR 97331, USA. E. D. Maloney, Department of Atmospheric Science, Colorado State University, 1371 Campus Delivery, Fort Collins, CO , USA. (emaloney@atmos.colostate.edu) 5of5