Baroclinic waves in the northern hemisphere of Mars as observed by the MRO Mars Climate Sounder and the MGS Thermal Emission Spectrometer

Abstract

The climatology of baroclinic waves in the northern hemisphere of Mars is investigated through analysis of observations by the infrared sounders on Mars Reconnaissance Orbiter (MRO) and Mars Global Surveyor (MGS). We focus on the lowest scale height above the surface, where the waves have a large impact on the Martian dust cycle. Profiles retrieved by the MRO Mars Climate Sounder (MCS) rarely reach the lower atmosphere at the season and location of interest. To fill this gap, we turn to observations in the MCS B1 channel (32 microns) when the instrument is viewing the surface. The signature of baroclinic waves appears in these data because of dust-related emission from the lower atmosphere and wave-induced variations of surface temperature. We supplement the MCS data with measurements of temperature at the 610-Pa pressure level from the MGS Thermal Emission Spectrometer (TES). Both data sets provide systematic coverage in latitude and longitude at two local times. Characteristics of baroclinic waves are derived through analysis of observations with a combined duration of about 8 Mars years. Basic results include least-squares solutions for wave amplitude and period at zonal wavenumber 1-3; the resolution is 4° in latitude and 14 solar days in time of observation. There is a strong similarity between the baroclinic waves observed by MCS and TES, which confirms the sensitivity of the MCS B1 channel to wave activity at pressures near 610 Pa. In all 8 Mars years, the baroclinic waves exhibit periodic transitions among modes with different zonal wavenumbers and a distinctive solstitial pause. Although the weather in each Mars year is unique in some respects, a composite of results from all years reveals a well-defined wave climatology. At each zonal wavenumber, large amplitudes are restricted to a pair of seasonal windows positioned symmetrically about the winter solstice. The wave-2 mode is strongest in early autumn and near the vernal equinox, whereas wave 3 is the dominant mode in mid-autumn and mid-winter, immediately before and after the solstitial pause. The interaction between baroclinic waves and dust storms is investigated through comparisons with spacecraft measurements of dust opacity. A strong wave-3 mode is often present during the initial growth phase of large, seasonal dust storms, which reflects the importance of wave-generated frontal dust storms in triggering these events. The wave-3 amplitude then decreases rapidly as the dust storm evolves; this occurs routinely in all Mars years considered here in connection with both mid-autumn "A" storms and mid-winter "C" storms. In some years A-storm suppression of the wave-3 mode marks the beginning of the solstitial pause. These results provide a basis for testing and development of Mars General Circulation Models as well as context for interpreting contemporaneous observations, such as spacecraft images of frontal and flushing dust storms.

Keywords: Atmospheres; Infrared observations; Mars; atmosphere; climate; dynamics.

Figure 1:

Lower abscissa: Vertical structure of baroclinic waves derived from RO temperature profiles (Hinson, 2006; Hinson and Wang, 2010). Results are shown for (orange) a wave-1 mode with a period of 6.9 sols at 66°N in autumn (L_s =204–212°) of MY27; (light blue) a wave-2 mode with a period of 3.0 sols at 69°N in autumn (L_s =190–200°) of MY26; and (dark blue) a wave-3 mode with a period of 2.3 sols at 64°N in winter (L_s = 316–334°) of MY25. Upper abscissa: (black) fraction of MCS profiles that reaches a given pressure level at the season and location where baroclinic waves appear. Dots mark the standard pressure levels in the MCS retrievals.

Figure 2:

Spatial coverage of MCS surface observations from 13 consecutive orbits (about 1 sol) at L_s =25° of MY30. The local time is about 15 h for observations on the dayside (light blue) and about 3 h for observations on the nightside (dark blue).

Figure 3:

Samples of 32-micron brightness temperature for the observations in Fig. 2. Dayside (light blue) and nightside (dark blue) temperatures converge at the pole but differ substantially at mid-to-low latitudes. These measurements are from early spring (L_s = 25°) of MY30.

Figure 4:

Least-squares spectrum showing results from a single bin centered at 48°N and L_s = 226° of MY29. The spectrum is dominated by a wave-3 mode with an amplitude of 4.5 K and a frequency of 0.47 sol⁻¹ (a period of 2.1 sols). The frequency range is appropriate for satellite observations twice per orbit (Salby, 1982). The spectrum is oversampled for clarity. Secondary peaks offset by ±0.107 sol⁻¹ from the primary peak result from use of a rectangular window function.

Figure 5:

Data from Fig. 4 as viewed in the longitude-time domain. The horizontal axis is longitude in a frame moving eastward at the zonal phase speed of the wave (56° sol⁻¹). Light and dark blue dots denote data from the dayside and nightside, respectively. The black line is the the least-squares fit to the data.

Figure 6:

RSS amplitude of baroclinic waves at zonal wavenumber 1–3. Each panel shows 1 MY of measurements beginning at the summer solstice (L_s =90°). The label at the lower left is the MY at the winter solstice (L_s =270°). TES results are shown in (A) and (B); MCS results are shown in (C)–(F). White bands denote data gaps. Panels (C)–(F) show the 150-K contour (white) of zonal-mean, daytime surface temperature, which tracks the edge of the CO₂ ice cap. Note that different color bars are used for results from TES and MCS.

Figure 7:

Variations of MCS noise σ_n with latitude and season in MY30 (color). White lines show contours of zonal-mean, daytime surface temperature. Contours range from 150 K to 240 K with a spacing of 15 K. The arrow indicates the midpoint of the largest frontal dust storm in MY30 (Wang and Richardson, 2015), which coincides with a notable enhancement of σ_n.

Figure 8:

MCS results at zonal wavenumber 2 in MY30, showing the wave period (A) in all bins where the SNR exceeds 0.35, and (B) in bins excluded from (A). The color bar is logarithmic. The black line in (A) separates atmospheric waves (to the north) from measurement noise (to the south). The white line in (A) is the 150-K contour of zonal-mean, daytime surface temperature, which tracks the edge of the CO₂ ice cap.

Figure 9:

TES results at zonal wavenumber 2 in MY26, showing the wave period (A) in all bins where the amplitude exceeds 1.7 K, and (B) in bins excluded from (A). The format is the same as in Fig. 8. Gray shading indicates data gaps.

Figure 10:

(A)–(D) Histograms of wave period for TES observations where the amplitude exceeds 1.7 K. This includes all TES results except those from L_s =180–270° of MY25, which are affected by a PEDS. (E)–(L) Corresponding results from all MCS observations (MY28–33) where the SNR exceeds 0.35. MCS results are shown separately for observations (E)–(H) to the north and (I)–(L) to the south of the latitude boundaries discussed in Section 5.2.1. A period of 16/3 sols is marked by a vertical white line in (J)–(L). A different vertical axis is used in each row.

Figure 11:

(A)-(C) MCS measurements of wave amplitude in MY30. Results are shown only in bins where the SNR exceeds 0.35. The black line in each panel is an empirically derived boundary (defined in Section 5.2.1) that separates reliable measurements of baroclinic waves at mid-to-high latitudes from surface-related noise at low latitudes. (D)-(F) TES measurements of wave amplitude in MY26. Results are shown only in bins where the amplitude exceeds 1.7 K. Gray shading indicates data gaps. Fig. 12 shows the corresponding measurements of wave period.

Figure 12:

(A)-(C) MCS measurements of wave period in MY30. Results are shown only in bins where the SNR exceeds 0.35. The black line in each panel is an empirically derived boundary (defined in Section 5.2.1) that separates reliable measurements of baroclinic waves at mid-to-high latitudes from surface-related noise at low latitudes. (D)-(F) TES measurements of wave period in MY26. Results are shown only in bins where the amplitude exceeds 1.7 K. Gray shading indicates data gaps. The color bar is logarithmic rather than linear so that the colors are distributed more evenly among the three zonal wavenumbers. Fig. 11 shows the corresponding measurements of wave amplitude.

Figure 13:

Seasonal evolution of the average amplitude at (orange) wave 1, (light blue) wave 2, and (dark blue) wave 3 in (A) MY24–26, (B) MY28–30, and (C) MY31–33. Gray bands denote data gaps. The horizontal bar in (A) identifies a period when wave activity was strongly affected by a PEDS. As in Figs. 6 and 11, different amplitude scales are used for TES and MCS. See Appendix B for a link to the data shown in this figure.

Figure 14:

The meridional average of wave amplitude measured by MCS at (orange) wave 1, (light blue) wave 2, and (dark blue) wave3. For clarity, separate panels are used for (A) waves 2 and 3, and (B) wave 1. This figure includes all results from the time period covered by Fig. 13B and 13C. See Appendix B for a link to the data shown in this figure.

Figure 15:

The meridional average of wave amplitude measured by TES at (orange) wave 1, (light blue) wave 2, and (dark blue) wave 3. For clarity, separate panels are used for (A) waves 2 and 3, and (B) wave 1. Observations affected by the PEDS that occurred in autumn of MY25 have been excluded from this figure. The results from each MY are shown individually in Fig. 13A. See Appendix B for a link to the data shown in this figure.

Figure 16:

The meridional average of wave period measured by (A) TES and (B) MCS at (orange) wave 1, (light blue) wave 2, and (dark blue) wave 3. Different symbols are used to distinguish the strongest mode in each bin of L_s (large circles) from the weaker modes (small triangles). This figure shows results from the same time periods covered by Figs. 14 and 15. TES results affected by the PEDS in autumn of MY25 have been excluded. See Appendix B for a link to the data shown in this figure.

Figure 17:

(color) Maps of zonal-mean dust column absorption at 9.3 microns (Montabone et al., 2015) versus latitude (left ordinate) and L_s. Black contours show zonal-mean opacities at levels denoted by tick marks in the color bar, which applies to all seven panels. A contour map of wave-3 amplitude is superimposed (yellow), showing the seasonal evolution of the wave and its location in latitude; the contour spacing is 1 K for TES (top row) and 0.5 K for MCS (all other rows). Line drawings show the average amplitude (right ordinate) of the wave-3 mode (solid white line in all panels) and the wave-2 mode (dashed white line in MY31 only).

Figure 18:

(A)–(C) Wave amplitude derived from measurements of T₆₁₀ spanning 1 MY (L_s = 90° of MY30 through L_s =90° of MY31). (D)–(F) Analogous results derived from contemporaneous measurements of brightness temperature in the MCS B1 channel. See caption to Fig. 11 for further explanation.

Figure 19:

(A)–(C) Wave period derived from measurements of T₆₁₀ spanning 1 MY (L_s =90° of MY30 through L_s =90° of MY31). (D)–(F) Analogous results derived from contemporaneous measurements of brightness temperature in the MCS B1 channel. Arrows identify specific modes discussed in the text. See caption to Fig. 12 for further explanation.