The diverse meteorology of Jezero crater over the first 250 sols of Perseverance on Mars

The study investigates the near-surface meteorology at Jezero crater during the first 250 sols of NASA’s Perseverance rover mission, aiming to characterize the processes governing the atmospheric surface layer (ASL) on Mars. Leveraging the Mars Environmental Dynamics Analyzer (MEDA), which provides the most complete environmental station sent to another planet, the work targets diurnal and seasonal cycles of temperature, pressure, winds, aerosols, and humidity. The context includes Jezero’s topographic setting near the Isidis Planitia basin and spring-to-early-summer season (Ls = 6°–121°). The purpose is to quantify the surface energy budget (SEB), vertical thermal structure, wind regimes, atmospheric dust and clouds, and hydrologic cycle to improve models, support rover operations (including Ingenuity flights and sample return planning), and understand interactions between local surface properties (albedo, thermal inertia) and atmospheric dynamics. The results address how local and regional phenomena—convection, slope flows, dust devils, clouds, gravity waves, and thermal tides—shape the environment at Jezero.

The paper builds on a long heritage of Martian meteorology from Viking landers, Pathfinder, Phoenix, Mars Exploration Rovers (Mini-TES), Curiosity’s REMS, and InSight, as well as boundary layer reviews and Mars Climate Database modeling. Prior studies identified ASL behavior, turbulence spectra, thermal tides, dust devils, and seasonal dust/cloud climatology. Observations at Gale and Elysium Planitia documented gravity waves, dust storms, and surface energy budgets; Viking and subsequent orbiter/rover datasets established diurnal opacity changes and aphelion cloud belt properties. The present work extends these by providing the first in situ SEB on Mars and continuous, multi-height temperature and aerosol observations at Jezero, enabling refined comparisons with single-column and large-eddy simulations and highlighting local topographic and surface property effects beyond one-dimensional expectations.

Instrumentation and operations: MEDA acquires autonomous sessions typically covering >50% of a sol, alternating hour-long sessions every other hour to sample full diurnal cycles every two sols. Components include: Thermal Infrared Sensor (TIRS) with channels for downwelling IR (IR1), air temperature (IR2), reflected shortwave (IR3), upwelling longwave (IR4), and ground temperature (IR5); Radiation and Dust Sensor (RDS) comprising upward and lateral photodiodes (190–1200 nm) and SkyCam imager; Atmospheric Temperature Sensors (ATS) at 0.85 m (two sensors) and 1.45 m (three sensors around the mast) and TIRS-derived temperature at ~40 m; Pressure Sensor (PS) with two capacitive transducers; Humidity Sensor (HS) with continuous and high-resolution interval modes; wind sensors on dual booms at ~1.5 m.

• Data selection and corrections: ATS selection uses wind orientation to choose downwind sensors minimizing rover thermal perturbations; RTG effects identified and excluded when present; rover geometry accounted for in wind retrieval via CFD-based weighting of boom data.

Surface energy budget (SEB): SEB is computed as G = (SW↓ − SW↑) + (LW↓ − LW↑) + Tr − Lr, with radiative fluxes measured by RDS/TIRS extended to full shortwave (0.2–5 μm) using COMIMART radiative transfer and to longwave (5–80 μm) using UH/FMI single-column model (SCM) with emissivity ε=0.99. Latent heat term Lr is neglected (no detected surface ice formation/sublimation). Turbulent heat flux Tr is estimated via bulk drag-transfer method using measured U and Ta at 1.45 m and stability corrections (von Kármán constant k=0.4, roughness z0=1 cm, Richardson-number-dependent function).

Thermal inertia (TI) and albedo: TI is retrieved by solving the 1-D heat conduction equation with MEDA SEB as upper boundary and fitting diurnal ground temperature amplitude and minimum; soil properties assume ρc=1.2×10^6 J m^−3 K^−1 and deep boundary at 3 e-folding depths. Broadband albedo (0.3–3.0 μm) is computed from TIRS SW↑ and COMIMART-extended RDS SW↓.

Aerosol optical depth (OD) and particle properties: SkyCam derives OD (550–800 nm) from direct solar imaging with calibration to MastCam-Z. RDS multi-wavelength/geometry observations, combined with radiative transfer (including T-matrix for non-sphericity), retrieve dust OD and effective radius by fitting simulated to observed sky radiance; lateral photodiodes constrain phase function near Sun; top photodiodes constrain spectral dependence. TIRS IR1/IR2 with EMIRS temperature profiles retrieve thermal IR aerosol OD across all local times (uncertainty ±0.03).

Humidity and pressure analysis: HS provides RH and sensor temperature; with PS, night-time VMR is derived. Spectral analyses of temperature, pressure, and wind fluctuations quantify turbulence regimes; Fourier decomposition of pressure and temperature time series isolates diurnal, semidiurnal, and higher-order tidal components; detrending reveals gravity waves and multi-sol baroclinic signals.

• Surface energy and surface properties:

  • First in situ SEB on Mars measured; single-column models and MEDA agree well on diurnal fluxes.
  • Thermal inertia varied substantially along the traverse, ranging roughly 180–605 (J m−2 K−1 s−1/2), influencing night-time inversion strength; best-fit examples: TI ≈ 230 (sol 30), 605 (sol 125), 290 (sol 209).
  • Albedo exhibits non-Lambertian behavior with minima near noon and increases at large solar zenith angles; specular effects observed when SZA ≈ 55°.
  • RTG heating within TIRS field of view is negligible for ground temperature (<0.5 K).

• Vertical thermal structure and turbulence:

  • Simultaneous temperatures at surface, 0.85 m, 1.45 m, and ~40 m resolve ASL regimes: daytime convection, evening transition, night-time stability, and morning transition.
  • Daytime unstable gradients reach dT/dz ≈ −35 K m−1 near noon; night-time stable gradients peak around +8 K m−1 in the first meter; adiabatic gradient is ~−0.0047 K m−1.
  • Temperature fluctuations rise after sunrise, peak around local noon with amplitudes up to ~10 K, subside near sunset, and re-intensify during night-time inversion breakup.
  • Power spectral densities during convective hours show turbulence-like slopes; distinct forced/inertial/dissipative regimes identified at multiple heights.

• Winds, dust devils, and aeolian activity:

  • Mean wind speeds show diurnal cycle: maxima ~7 m s−1 in afternoon, minimal near 04:00–06:00 LTST; gusts up to 25 m s−1 at midday; nocturnal fluctuations ~2–4 m s−1 and 5–7 m s−1 during convection.
  • Diurnal directionality dominated by upslope (day) and downslope (night) flows shaped by local topography; daytime upslope from roughly SE with night-time reversal.
  • Jezero exhibits very high dust devil (DD) activity: pressure drops ~0.3–6.5 Pa, durations 1–200 s; where winds co-observed, diameters ~5–135 m and rotational speeds −4 to 24 m s−1; some events caused measurable albedo changes in radiance ratios.

• Aerosols, clouds, and opacity:

  • SkyCam OD reveals persistent morning-afternoon asymmetry during clear season: morning OD ~0.5 vs afternoon ~0.4 (550–800 nm); aphelion cloud belt likely contributes to afternoon opacity and morning–afternoon differences.
  • RDS retrievals indicate dust effective radii ~1.2–1.4 μm; non-sphericity accounted for with T-matrix.
  • Cloud signatures detected daytime and twilight; color-index analyses and radiative transfer place cloud tops around or above ~40–45 km; particle sizes >1 μm consistent with water-ice.
  • TIRS thermal IR OD retrievals across all local times show more night-time opacity early (sol 30, Ls ~20°) and strong diurnal variation near aphelion peak (sol 200) with maximum shortly after dawn.

• Humidity and hydrologic cycle:

  • Early-morning RH maxima typically ~15–30% (relative to HS temperature), with absolute maximum observed ~29% RH; maximum VMR tends to occur around midnight.
  • Night-time VMR shows a seasonal minimum near Ls ≈ 70°, then increases with variability later in season.
  • A notable VMR increase occurred evening of sol 104, with concurrent cooling; frost conditions were possible early sol 105 when frost point (1.45 m) and ground temperature converged.

• Pressure, tides, and waves:

  • Seasonal mean pressure rose from ~735 Pa (sols 15–20, Ls 13°–16°) to ~761 Pa (sols 99–110, Ls 52°–57°), then declined to ~650 Pa by sol 250 (Ls 125°).
  • Multi-sol oscillations (3–5 sol periods) with amplitudes ~1–3 Pa indicate high-frequency traveling baroclinic waves.
  • Intraday gravity-wave-like oscillations show peak-to-peak ~0.2–0.4 Pa with 12–20 min periods.
  • Thermal tides strongly modulate pressure: up to six Fourier components; maximum relative change (δP/P)max ≈ 0.013; semidiurnal component shows a pronounced decrease between sols 20 and 50, possibly linked to dust loading or polar cap edge disturbances.
  • Temperature tides at 1.45 m exhibit half-amplitudes of ~26 K (diurnal), 2–6 K (semidiurnal), and 2–4 K (terdiurnal).

The observations demonstrate that Jezero’s near-surface environment is governed by a complex interplay of local and regional processes. The first in situ SEB quantification validates radiative–thermal modeling while revealing non-Lambertian surface reflectance, critical for interpreting orbital albedo variations and surface property retrievals. Multi-level temperature measurements capture canonical ASL regimes and show that nocturnal stability and inversion breakup can be strongly modulated by local thermal inertia, highlighting lateral heterogeneity that one-dimensional radiative equilibrium models may not capture. The wind climatology corroborates predictions of slope-driven daytime upslope and night-time downslope flows, with additional modulation by regional circulation and the Hadley cell—conditions that drive active aeolian processes and abundant dust devils that alter local surface albedo and dust loading. Continuous opacity monitoring confirms pronounced morning–afternoon differences and widespread dawn clouds during the aphelion season, offering a new diurnal perspective on dust–ice interactions previously limited by local-time sampling. The humidity record reveals a more intricate nocturnal hydrologic cycle than models predict, including events suggesting advection of distinct air masses and potential near-surface frost conditions, with implications for regolith–atmosphere exchange and near-surface photochemistry. Pressure spectra and Fourier components show gravity waves and multi-sol baroclinic activity during a nominally low-wave season, and thermal tides with amplitudes and timing that differ from Gale and Viking, likely due to basin-scale topography and basin–exterior exchanges. Collectively, these findings refine our understanding of Martian boundary-layer dynamics, aerosol–radiation interactions, and the hydrologic cycle at Jezero, informing operations (e.g., flight planning) and environmental constraints for sample caching and potential sample return.

This study presents the first comprehensive, multi-parameter, in situ characterization of Jezero crater’s ASL over 250 sols, including the first direct SEB measurement on Mars and continuous multi-height thermal, wind, aerosol, and humidity observations. Key contributions include: quantifying the variability of thermal inertia and non-Lambertian albedo; documenting strong diurnal thermal gradients and turbulence regimes; confirming slope-dominated wind patterns with frequent and energetic dust devils; resolving diurnal aerosol/cloud cycles with dawn enhancements and cloud tops above ~40 km; revealing a complex nocturnal humidity cycle with episodic events and potential frost conditions; and identifying gravity waves and multi-sol baroclinic oscillations alongside variable thermal tides. These results emphasize the role of local surface properties and topography in shaping near-surface meteorology and provide essential environmental context for Perseverance operations and Mars Sample Return planning. Future work should expand to dusty-season observations, integrate direct turbulence measurements (including w′), refine terrain-coupled mesoscale and LES modeling, and further investigate regolith–atmosphere water exchange, cloud microphysics, and topography–tide–wave interactions.

• Temporal sampling: MEDA runs hour-long sessions alternating every other hour; while coverage exceeds 50% of each sol, some diurnal windows are interpolated every two sols.
• Instrument perturbations: Night-time ATS at 1.45 m can be affected by RTG thermal contamination under certain wind orientations; procedures mitigate but cannot entirely eliminate such effects. Some pressure drops at night were linked to RTG thermal pulses.
• Turbulent flux estimation: Lack of direct vertical wind (w′) measurements requires use of bulk drag-transfer formulations for turbulent heat flux, introducing stability and roughness-related uncertainties.
• Spatial heterogeneity: TI and albedo retrievals assume locally homogeneous terrain; rover traverses across heterogeneous surfaces can complicate attribution of air temperature responses.
• Retrieval uncertainties: TIRS aerosol OD uncertainty ~±0.03; SkyCam opacity uncertainties ~0.07 (AM–PM differences ~0.08). Albedo relative error <10% near noon and <20% near sunrise/sunset.
• Seasonal context: Observations focus on northern spring to early summer (aphelion season) with relatively low dust loading; generalization to dusty seasons requires further data.
• Incomplete affiliation mapping and cross-instrument comparisons are beyond the scope of this dataset.