Distinct Mineralogy and Age of Individual Lava Flows in Atla Regio, Venus Derived From Magellan Radar Emissivity

Abstract

NASA’s Magellan mission revealed that many Venus highlands exhibit low radar emissivity values at higher altitudes. This phenomenon is ascribed to the presence of minerals having high dielectric constants, produced or stabilized by temperature-dependent chemical weathering between the rocks and the atmosphere. Some large volcanoes on Venus have multiple reductions of radar emissivity at varying altitudes. The authors present morphological maps of major lava flow units at Maat, Ozza, and Sapas montes and compare them to radar emissivity. Sapas has a single reduction in emissivity values at 6,054.6 km, while Maat and Ozza have several reductions at altitudes of 6,052.5–6,056.7 km. Emissivity values are highly spatially correlated to individual lava flows indicating that minerals in the rocks control the emissivity signature. The emissivity patterns at these volcanoes require at least four individual ferroelectric mineral compositions in the rocks that are highly conductive at Curie temperatures of 693–731 K. These temperatures are compatible with chlorapatite and some perovskite oxides. Modeling the minimum volumes of ferroelectrics (10–100s ppm) shows the volume and type of ferroelectric may vary over the lifetime of a single volcano. The modeled volumes of ferroelectrics in Ozza and Sapas are greater than in Maat, consistent with the production of ferroelectrics via weathering over a longer period of time, and supporting the idea that Maat has younger volcanic activity. The stratigraphic relationship of Maat’s youngest flows with impact craters may indicate the timeframe of the production of specific ferroelectrics via chemical weathering is over 9–60 Ma.

Plain Language Summary

NASA’s Magellan mission showed that most Venus summits have low values of radar emissivity at higher altitudes. These “anomalies” are due to conductive minerals produced or stabilized through chemical weathering reactions between the rocks and the atmosphere, which depend on rock and atmospheric composition, and the degree of weathering (surface age). While most highlands display an emissivity anomaly above a single altitude, the authors examine large volcanoes on Venus that have multiple anomalies at varying altitudes. The authors map the lava flow units on volcanoes in Atla Regio (Sapas, Maat, and Ozza montes), and look at their radar emissivity values over time. Sapas has a single reduction in emissivity, while Maat and Ozza have several reductions even at high temperatures. Emissivity anomalies are spatially correlated to individual lava flows, indicating that rock mineralogy controls the emissivity signature. The authors find that the type and amount of these minerals differs from one volcano to another and can change over the lifetime of an individual volcano, indicating differences in lava composition and exposure age. The composition and stratigraphy of Maat flows suggest that the edifice is the youngest of the three volcanoes and it last erupted less than 60 million years ago.

1. Introduction

Previous studies of NASA’s Magellan radar data showed that most of the Venus highlands exhibit a single reduction in radar emissivity values at high altitudes, above 6,053 km (Klose et al., 1992; Pettengill et al., 1992). This decline in radar emissivity with altitude is ascribed to the presence of minerals with a high dielectric constant produced by chemical weathering reactions between the rocks and the near-surface atmosphere. It is expected from theory that materials with high dielectric constants will enhance their radar reflectivity and lower their radar emissivity (Campbell, 1994; Pettengill et al., 1992). Proposed minerals include: (1) pyrite produced through sulfidation and/or oxidation of iron (Berger et al., 2019; Klose et al., 1992; Kohler, 2016; Pettengill et al., 1988; Port et al., 2016; Sempich et al., 2020; Wood & Brett, 1997), (2) coatings formed by condensation onto the rock as “metallic frosts” like lead and bismuth sulfides (Brackett et al., 1995; Kohler et al., 2015; Pettengill et al., 1996; Port et al., 2020; Schaefer & Fegley, 2004), and (3) ferroelectrics (e.g., perovskite or chlorapatite) that become highly conductive at certain temperatures (Arvidson et al., 1994; Shepard et al., 1994; Treiman et al., 2016). These reactions are a function of rock composition, atmospheric composition, temperature, and degree of weathering or surface age.

Recent global analysis of the emissivity of all major Venus highlands (Brossier, Gilmore, & Toner, 2020) reveals that the two largest volcanoes on Venus, Maat and Ozza Montes, exhibit multiple reductions in radar emissivity at a broad range of altitudes including, atypically, lowlands (the term “lowlands” is used throughout this manuscript and corresponds to altitudes below 6,053 km, which comprise the lowlands and upland rolling plains of Masursky et al. (1980)). This phenomenon that was also reported in earlier studies (e.g., Robinson & Wood, 1993; Wilt, 1992). These patterns indicate differences in the volume and/or composition of minerals with a high dielectric constant in the volcanic system. The low-altitude anomalies also require the formation of minerals with a high dielectric constant at higher temperatures than observed at other Venus highlands. To better understand these patterns, the authors perform detailed mapping of these volcanoes, examine the correlation of mappable geomorphic units to radar emissivity and describe variability in emissivity with location and stratigraphic position (time). The authors determine the altitudes and corresponding temperatures of the emissivity anomalies and model the type and volume of candidate high dielectric minerals that could explain the observations. Changes in the nature of the emissivity anomalies may yield insight into the age and compositional evolution of the volcanic system.

Here, the authors focus our investigation on Sapas, Maat, and Ozza montes, three large volcanoes in Atla Regio (Figure 1). Atla Regio is a volcanic region where five rifts intersect (Dali, Ganis, Parga, and two other chasmata), facilitating rift-associated volcanism in the form of large volcanic edifices, such as Maat and Ozza in Atla Regio, Theia in Beta Regio, and Yunya-mana in Phoebe Regio (Senske et al., 1992). The central part of Atla Regio is dominated by Ozza, whereas Maat is located on the flank of the rift system (Senske et al., 1992). Sapas is geographically separated from the Atla Regio rift zones, as it is located in volcanic plains of Rusalka Planitia (Figure 1). Brossier, Gilmore, and Toner (2020) report that Maat and Ozza have several emissivity excursions at different altitudes, while Sapas only has a single emissivity excursion. Sapas is therefore included for comparison purposes. The authors also consider Ningyo Fluctus, a flow field that originates on the southern flank of Ozza, and exhibits similar emissivity excursions. Figure 2 is a map of Magellan emissivity data overlapping radar backscatter images that cover the three volcanoes. Blue arrows correspond to low emissivity excursions and red arrows are high emissivity summits.

Figure 1.

SAR map of Atla Regio (198°E, 2°N): Sapas, Maat, and Ozza montes, and a large flow field (Ningyo Fluctus) that is possibly associated with Ozza’s activity. Peak heights of the three volcanoes are indicated below their names. The map has a simple cylindrical projection and north is up. SAR, Synthetic Aperture Radar.

Figure 2.

Magellan emissivity maps overlapping SAR images of the three volcanoes: Sapas (188°E, 8°N), Maat, and Ozza (198°E, 4°N). Ningyo Fluctus (206°E, 5°S), a flow field located near Atla Regio and Parga Chasma. Magellan emissivity and SAR images covering the study area are provided in the online repository linked to this work (Brossier, Gilmore, Toner, & Stein, 2020). SAR, Synthetic Aperture Radar.

This paper is structured in the following way. The authors first map the flow units that compose the three volcanoes (Section 3.1) and correlate them with emissivity and elevation (Section 3.2). The authors then apply a model using the emissivity excursions to estimate the volume of minerals with high dielectric constants and consider potential minerals that could be responsible for the excursions observed (Section 4). Finally, the authors discuss the diverse mineralogy of the lava flows (Section 5) and the changes that the volcanic systems underwent through time (Section 6).

2. Data & Methods

The radar properties of Sapas, Maat, and Ozza montes are extracted using datasets collected during the Magellan mission (frequency = 2.4 GHz, λ = 12.6 cm). The authors identified the major lava flows of the three volcanoes with the Cycle 1 left-looking Synthetic Aperture Radar (SAR) images (FMAPS, 75 m.px⁻¹). The authors derived altimetry and emissivity from the Magellan global topography data records (GTDR) and global emissivity data records (GEDR). Altimetry data have a spatial resolution ranging from ~10 km at periapsis (ca. 10°N latitude) to ~20 km near the poles (ca. 90°N and 70°S) when the orbiting spacecraft was high above the planet. Emissivity data were collected while the spacecraft was operating in radiometer mode. The spatial resolution of the emissivity data varies from ~20 km near periapsis to ~80 km at high was high latitudes (Pettengill et al., 1991). Near-global mosaics are produced in the GTDR and GEDR data products that are publicly available through the USGS websites (see Data Availability Statement). The two mosaics are resampled to a spatial resolution of 4.6 km.px⁻¹ (scale of 22.7 px.deg⁻¹). Heights given throughout this paper may differ from those found in the past literature as our values are measured from a mean planetary radius (MPR) taken as 6,051.80 km (Ford & Pettengill, 1992). Altimetry and emissivity data are extracted from these mosaics to return scatterplots of the variation of emissivity with altitude for each region, as in Brossier, Gilmore, and Toner (2020) and Brossier and Gilmore (2021).

Selection and extraction processes are performed using the ArcGIS 10.6 (ESRI) software package, while the plots are produced with RStudio software. The authors also calculate temperatures as a function of elevation using the Vega 2 lander data (Lorenz et al., 2018; Seiff, 1987). To date, the Vega 2 lander is the only probe that has returned in-situ measurements of the temperature below 12 km down to the surface. Brossier, Gilmore, and Toner (2020) calculated a regression for these data that provide a reasonable linear relation between temperature and altitude. These temperatures are essential to test the stability and reactivity of candidate minerals under Venus conditions.

3. Sapas, Maat, and Ozza Montes

3.1. Definition of Flow Units

Volcanic flow units are defined based on radar brightness and morphology. The correlation of a flow associated with an individual eruption is often difficult or impossible to singularly constrain in radar (Elachi et al., 1980; Gaddis et al., 1989). Thus throughout the present paper a mapped lava flow composite unit (hereafter “flow unit”) may refer to a feature likely composed by two or more flows that is the result of more than a single eruptive event. The brightness of a lava flow in SAR images is controlled by the combined effects of surface roughness at radar wavelength (12.6 cm for Magellan), the dielectric constant of the surface et al., 1980). At the Magellan incidence angle (~45° at Atla Regio for Cycle 1 data), and for emissivity values materials, the size and orientation of km-scale topographic slopes, and subsurface inhomogeneities (Elachi near the global average (~0.85), surface roughness dominates the return (Keddie & Head, 1994). Thus, the major cause of the difference between a bright and a dark flow is that the former is relatively rough at the cm-scale and the latter is relatively smooth (Keddie & Head, 1994, 1995). Low emissivity values (below 0.6–0.7) require a contribution from materials with a high dielectric constant (Klose et al., 1992; Robinson & Wood, 1993). The authors extracted DN values from the SAR images for each flow unit to support the 150 (~10 dB), while dark flows have DN values below 120 (~4 dB). Additionally, the spatial distribution distinction between radar-bright and radar-dark flows (Table 1). Bright flows usually have DN values above of variations in brightness within a flow may result in distinct textures. Since the authors are focusing on geologic context in the study of Sapas, Maat, and Ozza montes, the authors applied a strictly qualitative definition of texture. If the flows have moderate radar backscatter and vary little in brightness, the overall texture of the flows is smooth and relatively featureless. However, if radar brightness varies both within and between individual flows of a lava flow composite unit, this results in an overall mottled appearance with low to high radar backscatter. Thus, our units are defined by four major properties, following Keddie and Head (1994, 1995): (1) lava flow morphology (e.g., sinuous, narrow, broad, patchy), (2) overall flow brightness which is dominated by cm-scale roughness, (3) texture, and (4) spatial and inferred stratigraphic relationships (Table 1 and Figure S1). Closer views on the mapped flow units (morphology, brightness, texture, and spatial relationships) are illustrated by high resolution images in Figure S1.

Table 1.

Main Characteristics of the Mapped Flow Units (Figure 3)

Flow unit	Morphology	Radar backscatter	Texture	DN value	σ0 (dB)
Sapas Mons
S1	Broad, patchy	Moderate	Smooth	139 ± 11	7.5
S2	Narrow, sinuous	Low to high	Mottled	134 ± 11	6.6
S3	–	Low	Smooth	109 ± 14	1.5
S4	Narrow, sinuous	Low to high	Mottled	136 ± 15	7.0
S5	Patchy	Very high	Hummocky	163 ± 9	12.4
S6	–	Low to high	–	143 ± 19	8.4
Maat Mons
M1	Narrow, sinuous	Moderate	Smooth	141 ± 10	7.9
M2	Patchy	High	Hummocky	152 ± 12	10.3
M3	–	Low	–	108 ± 12	1.4
M4	–	Low	–	114 ± 11	2.7
M5	Broad, sinuous	Moderate	Smooth	139 ± 11	7.6
Ozza Mons
O1	Narrow, sinuous	Moderate	Smooth	132 ± 19	6.1
O2	Broad, patchy	Very high	Hummocky	166 ± 11	13.0
O3	–	Low to high	–	160 ± 20	11.8
O4	Patchy	Very high	Hummocky	173 ± 7	14.5
O5	–	High	–	160 ± 11	11.8
O6	Narrow, sinuous	Moderate	Smooth	148 ± 7	9.4
O7	Broad, sinuous	Low	Smooth	121 ± 11	4.1
O8	Broad, sinuous	Moderate	Smooth	147 ± 9	9.2

Notes: Terminology for morphology, backscatter and texture after Keddie and Head (1994) and Klose et al. (1992). DN values are extracted from SAR images and then converted to backscatter coefficient (σ0 in dB) after Campbell (1995).

Mapped lava flow units are defined for each volcano and placed in simple stratigraphic columns where lower numbered units are stratigraphically older (Figure 3). For Sapas Mons, the authors defined 6 flow units (Figure 3, top left). Flows of the S1 unit have a smooth texture with moderate radar backscatter and are exposed to the northwest and southwest of the volcano. The S2 unit exposures show a mottled texture with low to high radar backscatter and occurs in two large regions on the northwestern and southeastern flanks of the edifice. The S3 unit is characterized by radar dark materials exposed along its southeast flank. The S4 unit is similar in characteristics to the S2 unit; most flows are sinuous and vary in radar backscatter, resulting in a mottled texture for the unit as a whole. The majority of flows from the S4 unit are exposed on the southern flanks. In a few locations, some flows cross the underlying flows (S2–3 units) and clearly demonstrate the temporal relationship between units. The S5 unit stands out as the brightest part of Sapas with a hummocky texture, and its flows are slightly broader and less sinuous than those of the S2 and S4 units. The S5 unit flows are exposed almost entirely around the volcano summit. The S6 unit is defined by the two domes with scalloped margins (Guest et al., 1992) and associated radar dark deposits exposed of their flanks. The flow units identified at Sapas are consistent with the mapping provided in Keddie and Head (1994) (their units 1–6).

Figure 3.

Sketch maps of the lava flow units within the three volcanoes: (top-left) Sapas, (top-right) Maat, and (bottom) Ozza. Crosses (+) indicate the summit(s) of each volcano. Black box indicated an isolated patch of O2 unit discussed in text. The shapefiles (and auxiliary files) of each volcano are given in the online repository (Brossier, Gilmore, Toner, & Stein, 2020).

The lava flows at Maat Mons have been subdivided into 5 flow units (Figure 3, top right). The earliest M1 unit flows have a moderate radar backscatter and they extend west from the volcano. The M2 unit is defined by bright flows exposed on the southwestern flank of Maat. The M3 and M4 units include radar dark materials confined on the northern flank and the summit region, respectively. The flows of the most recent unit, M5, extend east from the central edifice. The flow units identified at Maat are in good agreement with preliminary mapping provided by Klose et al. (1992) and Keddie and Head (1994). A more detailed map of Maat summit (our M4 and M5 units) is provided in Mouginis-Mark (2016), where unit subdivisions are primarily based on morphology (e.g., digitate, sheet and filamentary flows). Our M4 unit comprises digitate and extensive sheet flows with a few boundaries faintly recognizable, while the M5 unit has narrow flows with filamentary (or braided) lobes at the margins (Mouginis-Mark, 2016).

This work includes the first detailed map of Ozza Mons. The lava flows at Ozza have been subdivided into 8 flow units (Figure 3, bottom). Flows of the O1 unit have a moderate radar backscatter and smooth texture and extend northwest from Ozza. The bright patchy flows from the O2 unit surround the volcano. The O2 unit includes a patch that is isolated from the rest of the mapped unit (black box in Figure 3) and is separated by the older O1 unit, resulting in a lateral discontinuity. This isolated patch appears to overlap and postdate the O1 unit, although its temporal relationship with the other flow units is uncertain. The visible similarity in both morphology and radar brightness support the idea that this flow patch belongs to the O2 unit. It could have originated during the same eruption events which produced the rest of the O2 unit (e.g., from a separated lateral vent). Flank eruption activity is likely to occur on Venus, as observed in Gula, Sif and Kunapipi montes (Stofan et al., 2001), and Idunn Mons (D’Incecco et al., 2017). The O3 unit lies within a tectonized caldera (Senske et al., 1992). The highest part of the O3 unit is a plateau with a very low radar return. The O4 unit is characterized by flows that are stratigraphically above the O3 unit and are relatively brighter. The O5 unit is a field of small shield volcanoes located near the summit plateau. The flows of the O6 unit extend north and northeast from the volcano. The O7 and O8 units comprise flows of a large flow field, Ningyo Fluctus, located near Ozza and at the margin of Parga rift zone. They have low and moderate radar backscatter, respectively, and both exhibit a smooth texture.

3.2. Emissivity Excursions

For each volcano, the authors define an emissivity excursion (e) as the region on an emissivity–elevation scatterplot (Figures 4 and 5) where radar emissivity values decrease and become distinct from the global plains average (~0.85). Emissivity and elevation values in Figures 4 and 5 are obtained for each flow unit as defined in the previous section (Figure 3), allowing us to identify any correlations between emissivity excursions and specific flow units within the three volcanoes.

Figure 4.

Elevation–emissivity plots obtained for (a) Sapas and (b) Maat montes (gray data points) and their flow units (color coded data points, see maps in Figure 3). Red lines in plots are reference values of emissivity at 0.7 (continuous) and 0.8 (dashed). Temperatures are given by the Vega 2 lander data (Lorenz et al., 2018; Seiff, 1987). Elevation (as planetary radius, in km) and emissivity values are reported as text files in the online repository (Brossier, Gilmore, Toner, & Stein, 2020).

Figure 5.

Elevation–emissivity plots obtained for Ozza Mons (gray data points) and its flow units (color coded data points, see maps in Figure 3). Annotations are the same as in Figure 4.

Most flow units at Sapas (S1–S5 units) participate in a single emissivity excursion (e1) and reach a minimum value at an elevation of ~6,054 km (Figure 4a). The bright flows from the S5 unit display a more gradual decline in emissivity above this altitude reaching an emissivity low of ~0.4 at 6,054.6 km, before returning to slightly higher emissivity values at the summit (S6 unit). The return to near-normal emissivity values seen on the S5 unit flows could be ascribed to compositional changes and/or the presence of fine-grained debris on the flanks of the domes (Keddie & Head, 1994). Fine grained materials are supported by a low radar return in SAR images in the S6 unit. Fine grained materials are predicted to result in low radar reflectivity and thus high radar emissivity (Campbell, 1994). The spatial resolution of the emissivity data prevents direct correlation of these fine-grained materials to emissivity since the emissivity footprint is an average of both S5 and S6 flow units (Palazzari et al., 1995).

At Maat, the bright flows from the M2 unit contain three distinct emissivity excursions at 6,056.2 km (e1), 6,055.4 km (e2), and 6,053.9 km (e3) (Figure 4b). The M1 unit is the main source for the excursion at 6,052.7 km (e4) and has another excursion at 6,053.6 km similar in elevation to the excursion at 6,053.9 km (e3). The radar dark materials identified on the northern flank (M3 unit) and in the summit region (M4 unit) of the volcano have generally high emissivity values (above 0.8). Flows of the M5 unit emanating from the summit also have high emissivity above 0.7 for the entire elevation range of Maat (6,052–6,061 km). However, they have decreases in emissivity at 6,055.4 km (e2) and 6,053.9 km (e3) that correspond to the elevations of the excursions in the M2 unit.

Finally, at Ozza, the tectonized summit unit (O3) and the O4 unit contain two emissivity excursions that reach values as low as 0.36 at elevations of 6,056.7 km (e1) and 6,055.7 km (e2), before returning to global average values at elevations above 6,057 km (i.e., the radar dark plateau) (Figure 5). Flows of the O1 unit display the same behavior but show a single primary emissivity excursion at 6,056.7 km (e1). The field of shield volcanoes found near the plateau (O5 unit) has an average emissivity of ~0.55 at elevations similar to previous units. The O2 unit is the main source for the sharp excursion to values of 0.53 at 6,052.5 km (e3) and also displays an emissivity excursion of ~0.4 at 6,055.7 km (e2). Flows of the O6 unit display a gradual decline in emissivity indistinguishable from those of the O2 unit, the O6 unit flows do not reach elevations as high. Both flow units in Ningyo Fluctus correspond to a low elevation excursion at 6,052.5 km (e3) despite their apparent roughness difference; the dark flow (O7 unit) is smoother than the bright one (O8 unit) based on backscatter values. This indicates that dielectric constant (not roughness) is dominating the emissivity signal.

4. Ferroelectric Model

Maat and Ozza lava flows display reductions in emissivity values with increasing elevation until a given altitude over which there is an abrupt return to higher emissivity (Brossier, Gilmore, & Toner, 2020; Pettengill et al., 1992; Wilt, 1992). This pattern is consistent with the behavior of ferroelectric minerals (Arvidson et al., 1994; Shepard et al., 1994; Treiman et al., 2016). Ferroelectric minerals undergo a phase transition and become very conductive at a specific temperature called the Curie temperature (T_c). The elevation of the emissivity excursion marks the Curie temperature for the minerals in the rocks. The temperature and altitude of the emissivity excursion are function of composition, while the magnitude of the excursion is a function of volume of the ferroelectric (Shepard et al., 1994). On Venus, this behavior has been reported in previous studies of Ovda Regio (Arvidson et al., 1994; Shepard et al., 1994; Treiman et al., 2016) and more recently on several volcanoes and coronae (Brossier, Gilmore, & Toner, 2020) and tesserae (Brossier & Gilmore, 2021). In Sapas, the pattern of the emissivity variation resembles that of ferroelectric behavior with only a slight upturn to higher emissivity values on the summit (Brossier, Gilmore, & Toner, 2020). However, this diffuse pattern can also be ascribed to fine-grained deposits that have flowed or fallen as debris from the domes (Keddie & Head, 1994). Thus, the authors can neither confirm nor rule out the ferroelectric behavior in Sapas, due to the resolution of the data (Palazzari et al., 1995).

If the authors assume that all observed excursions are due to ferroelectric minerals at their Curie temperatures, the authors can attempt to model their shapes to estimate the volume of ferroelectric mineral present in the rocks following Shepard et al. (1994). For a given emissivity value and radar incidence angle, the permittivity value (or dielectric constant) can be calculated using relations from Campbell (1995) (Equations 1–3) where E is the emissivity, Φ is the incidence angle converted from latitude (Table S1), ε is the permittivity for horizontally (E_h) and vertically (E_v) polarized radar signals. E_h approximates surfaces that are smooth at the radar while the average of E_h and E_v approximate rough surfaces.

$$E_h=(\sin 2\mathbf{\text{Φ}}\times \sin 2\theta )/{\sin }^2(\mathbf{\text{Φ}}+\theta ),$$ (1)

$$E_y=(\sin 2\mathbf{\text{Φ}}\times \sin 2\theta )/\left({\sin }^2(\mathbf{\text{Φ}}+\theta )\times {\cos }^2(\mathbf{\text{Φ}}-\theta )\right),$$ (2)

$$\theta ={\sin }^{-1}(\sin \mathbf{\text{Φ}}/\sqrt{\epsilon }).$$ (3)

The authors use the smooth and rough criteria to return the total span of modeled permittivity values for the three volcanoes. The smooth criterion (E_h) provides a minimum value while the rough criterion ([E_h + E_v]/2) gives a maximum value for the permittivity. This allows us to produce plots of the variations of permittivity with altitude for each volcano (Figure 5). Once the authors derive the permittivity, the authors can apply the ferroelectric model from Shepard et al. (1994) which assumes a material with ferroelectric inclusions which have an assumed Curie constant and model geometry (tabular, acicular, or spherical). The Curie constant is defined by the Curie-Weiss law (Equation 4) where ε_f is the permittivity of the ferroelectric, C is the Curie constant which is a function of the material, T is the temperature, and T_c is the Curie temperature (Rupprecht & Bell, 1964).

$$\epsilon f=C/\left(T-T_c\right)\text{for}T>T_c$$ (4)

Ferroelectrics exhibit Curie constants of the order of 10³ for di-hydrated salts like Rochelle Salt and Potassium Dihydrogen Phosphate (KDP), and 10⁵ for most perovskite oxides (Jona & Shirane, 1962). To date no Curie constant has been measured for chlorapatite. Here, the authors apply the model under specific conditions to give minimum volumes of ferroelectric minerals homogenously distributed in the volume of rock interrogated by Magellan radiometer (penetration depth of cm-m, e.g., Bondarenko et al., 2003). The authors model a Curie constant of 10⁵ and tabular shape for the inclusions, assumptions that are most consistent with rock-forming minerals. A change of the Curie constant to 10³ (Table 2) or acicular or spherical geometries will result in greater volumes of ferroelectrics than modeled here. The reader is referred to Shepard et al. (1994) for more details about the ferroelectric model.

Table 2.

Ferroelectric Modeling

Excursion (unit)	Min E	Altitude (km)	Temp. (K)	Max ε	Vol.a (ppm)	Vol.b (ppm)
Sapas Mons
e1 (S5)	0.395	6,054.6	7l2.2	36.7–68.3	45–87	4,470–8,720
Maat Mons
e1 (M2)	0.527	6,056.2	697.8	16.5–29.6	18–35	1,750–3,510
e2 (M2)	0.500	6,055.4	704.9	20.0–35.7	22–43	2,220–4,330
e3 (M2)	0.582	6,053.9	718.4	12.6–22.0	12–25	1,230–2,490
e4 (M1)	0.688	6,052.7	729.1	7.8–13.0	6–13	580–1,280
Ozza Mons
e1 (O3)	0.358	6,056.7	692.7	46.4–87.8	58–113	5,770–11,340
e2 (O3)	0.359	6,055.7	701.5	46.4–87.8	58–113	5,770–11,340
e3 (O2)	0.529	6,052.5	730.8	15.6–28.5	16–34	1,630–3,360

Notes: Maximum values permittivity (ε) calculated from minimum values of emissivity (E) after Campbell (1995) for smooth (lówer value) and rough (higher value) criterions, and volume estimated after Shepard et al. (l994) by using the Curie constants of (a) 10⁵ and (b) 10³.

All values associated with the emissivity excursions are reported in Table 2, including the lowest emissivity value observed, its corresponding altitude and temperature (Vega 2), highest derived permittivity (Campbell, 1995) and estimated volume of ferroelectric mineral (Shepard et al., 1994). Using the elevation and magnitude of the observed emissivity data, the authors generate model curves for each excursion based on Equation 3 in Shepard et al. (1994). The emissivity excursion (~0.4) observed on Sapas Mons reaches a minimum value at 6,054.6 km and corresponds to a Curie temperature of 712 K (Figure 4a). The derived permittivity at the excursion ranges from 37 (smooth criterion) to 68 (rough criterion) which corresponds to a minimum volume of ferroelectric mineral of 4–87 ppm (Figure 6). The multiple emissivity excursions (0.36–0.69) of Maat and Ozza montes reach their minima at altitudes ranging from 6,052.5 to 6,056.7 km, corresponding to Curie temperatures of 693–731 K (Figures 4b and 5). The highest permittivity values at these excursions vary from 8–88, while the minimum volumes of dielectric mineral range from 6 to 113 ppm (Figure 6).

Figure 6.

Elevation–permittivity plots obtained for the three volcanoes. Application of the ferroelectric model (Shepard et al., 1994) to reproduce the low emissivity excursions (color-coded curves). Elevation (as planetary radius, in km) and permittivity values used to reproduce the excursions are reported in the excel file provided in the online repository (Brossier, Gilmore, Toner, & Stein, 2020).

5. Diverse Mineralogy

In the volcanoes studied here, the authors commonly observe adjoining morphological units at the same elevation with different emissivity signatures. The authors also see a strong spatial correlation between the emissivity values and independently mapped lava flow units. For example, at Maat, the lobate boundaries of the M2 flows correspond almost exactly to low emissivity values, including fissure fed eruptions downslope from the main exposure of the unit (area around e3, Figure 7a). Similarly, at Ozza, materials in the O2 unit correspond to low emissivity values, visible in kipukas of M2 isolated by later flows from the O6 unit (Figure 7b). The close spatial association of emissivity excursions with independently mapped lava flow units indicates the emissivity signatures are contained in the rocks themselves, due to the presence of minerals (or compounds) with a high dielectric constant. The authors assume that the high dielectric minerals are intrinsic to the lava flows, in this case ferroelectric minerals that crystallize directly from the lava, or non-ferroelectric minerals are converted to ferroelectric minerals due to surface-atmosphere weathering over time. This is different than a “snow line” pattern seen in other regions of Venus (e.g., at Maxwell Montes, Treiman et al., 2016), where there is a sharp change in emissivity in all materials above a particular altitude suggesting a reaction largely controlled by temperature.

Figure 7.

Magellan emissivity maps overlapping SAR images of (a) the southwestern flank of Maat, and (b) the northern flank of Ozza. Maat’s M2 unit is correlated with the three emissivity excursions at 6,056.2 km (e1), 6,055.4 km (e2), and 6,053.9 km (e3). Ozza’s O2 unit matches with the low elevated excursion at 6,052.5 km (e3). This spatial association indicates that the emissivity anomalies are intrinsic to the flow units. SAR, Synthetic Aperture Radar.

The authors reiterate here that, for ferroelectric minerals, the elevation of the emissivity signature is related to the Curie temperature and the magnitude of the drop of emissivity is a function of the volume of the included minerals (Shepard et al., 1994). The diversity of transition temperatures and altitudes of the emissivity signatures observed in Maat, Ozza, and Sapas flows could be explained by several reasons: (1) each lava flow may contain a distinct mineralogy that has a different Curie temperature, and hence a transition occurring at a different elevation; (2) each lava flow may contain the same mineralogy but with subtle differences in composition, and hence varying the transition temperature and altitude; or (3) each lava flow has the same mineralogy, but experienced different atmospheric conditions during the surface weathering.

The flows of Ozza, Maat and Sapas require a minimum of four different ferroelectric minerals labeled A to D (Figure 6): (A) T_c = 693–698 K at Maat (e1) and Ozza (e1), (B) T_c = 702–705 K at Maat (e2) and Ozza (e2), (C) T_c = 712–718 K at Sapas (e1) and Maat (e3), and (D) T_c = 729–731 K at Maat (e4) and Ozza (e3). The high T_c materials are the dominant minerals in the flows of Ningyo Fluctus which emanate from the flank of Ozza (Figure 5).

Table 3 shows a list of known possible candidate ferroelectrics that have Curie temperatures in the range of (or close to) venusian surface temperatures (735 K at MPR, and 690 K at the summit of Maxwell Montes). This includes all minerals that have been described previously for Venus, plus additional minerals from a survey of the literature.

Table 3.

List of Ferroelectric Minerals With Curie Temperatures Near Venus Surface Temperature

Compound	Formula	Curie temp. Tc (K)	Curie constant	Refs	A	B	C	D
Barium Bismuth Tungstate	Ba(Bi0.7W0.3)O3	723	–	1	✗	✗	✗	✗
Cadmium Iron Niobate	Cd(Fe0.5Nb)O3	723	–	1	✗	✗	✗	✗
Cadmium Scandium Niobate	Cd(Sc0.5Nb0.5)O3	703	–	1	✗	✓	✗	✗
Chlorapatite*	Ca5(PO4)3Cl	675–775	–	2	✓	✓	✓	✓
Lead Bismuth Tantalate	PbBi2Ta2O9	703	–	1	✗	✓	✗	✗
Germanium Telluride*	GeTe	623–670	∼1.0 × 105	1,3,4	✗	✗	✗	✗
Lead Bismuth Tantalate–Niobate*	Pb2Bi(Ta,Nb)O6	693–748	–	1,5	✓	✓	✓	✓
Lead Tantalate–Niobate*	Pb(Ta,Nb)2O6	533–843	–	5,6	✓	✓	✓	✓
Lead Titanate*	(Pb(Ba,Sr,Ca))TiO3	110–763	1.1 × 105	1,5,6	✓	✓	✓	✓
Sodium Niobate-Tantalate*	Na(Nb,Ta)O3	627–753	–	1,5	✓	✓	✓	✓
Lutetium Chromite	LuCrO3	713	–	1	✗	✗	✓	✗
Potassium Lithium Niobate	K3Li2Nb5O15	703	–	1	✗	✓	✗	✗
Potassium Niobate–Tantalate*	K(Nb,Ta)O3	2–708	2.4 × 105	1,5,6	✓	✓	✗	✗
Strontium Bismuth Niobate	SrBi2Nb2O9	713	–	1	✗	✗	✓	✗
Rochelle Salt	NaKC4H4O6 · 4H2O	255–297	2.2 × 103	7	✗	✗	✗	✗
KDP	KH2PO4	123	3.3 × 103	8	✗	✗	✗	✗

Notes: Minerals proposed specifically for Venus are indicated by an asterisk (*). Rochelle salt and Potassium Dihydrogen Phosphate (KDP) are added for comparison. References: (1) Subbarao (1973); (2) Treiman et al. (2016); (3) Brackett et al. (1995); (4) Kadlec et al. (2011); (5) Shepard et al. (1994); (6) Young and Frederikse (1973); (7) Valasek (1927); (8) Busch and Scherrer (1935). Curie constants are given in Jona and Shirane (1962). Listed minerals can account ✓ (or not ✗) for emissivity excursions at (A) 693–698 K, (B) 702–705 K, (C) 712–718 K, and (D) 729–731 K.

Three ferroelectric minerals listed in Table 3 cannot account for any of the emissivity excursions seen in Atla Regio: the perovskites Barium Bismuth Tungstate, Cadmium Iron Niobate and the material Germanium Telluride (GeTe) which was proposed specifically for Venus by Brackett et al. (1995). Five minerals can explain all emissivity excursions observed in the three volcanoes, each of which has been proposed for Venus: chlorapatite (Treiman et al., 2016) and four perovskites: Sodium Niobate–Tantalate, Lead Titanate, Lead Tantalate–Niobate and Lead Bismuth Tantalate–Niobate (Shepard et al., 1994). Other minerals may correlate to a subset of the four observed excursions: for example, Lead Bismuth Tantalate can only account for the emissivity excursions in the highlands (cold temperatures). Conversely, Strontium Bismuth Niobate correlates only excursions in the lowlands (higher temperatures).

Perovskite oxides were proposed by Shepard et al. (1994) as candidate ferroelectric minerals for Venus. Perovskite-family oxides are a series of phases with the structure ABX₃, where X is typically oxygen. The mineral perovskite, CaTiO₃, is a natural occurrence of a wide range of compounds with this structure. Changes in the composition of the elements in the A and B site can change the Curie temperature and the conductivity of the ferroelectric compounds. For example, Shepard et al. (1994) report that minor change of the Pb abundance in lead perovskites can increase or decrease the Curie temperature (Rupprecht & Bell, 1964). More precisely, a 1% change in the Pb abundance changes the Curie temperature of about 8 K, corresponding to a 1 km change in the transition altitude. Similarly, the addition of rare earth elements to the structure of lead titanates may change both the Curie temperature and dielectric constant (e.g., Durán et al., 1989; Moure & Peña, 2015). The sensitivity of Curie temperature to small changes in composition must be considered in the assignment and origin of candidate ferroelectric minerals.

Treiman et al. (2016) suggested chlorapatite as a ferroelectric candidate for Venus. Apatite Ca₅(PO₄)₃ (OH, F, and Cl) is a minor but common mineral in igneous rocks, but on Earth it usually occurs as fluorapatite that is not ferroelectric. Primary fluorapatite might be altered to chlorapatite by chemical reaction with HCl from the deep (near surface) atmosphere of Venus, following Equation 5 (supplementary material from Treiman et al., 2016).

$${\text{Ca}}_5{\left({\text{PO}}_4\right)}_3{\text{F}}_{(\text{mineral})}+{\text{HCL}}_{(\text{gas})}={\text{Ca}}_5{\left({\text{PO}}_4\right)}_3{\text{CL}}_{(\text{mineral})}+{\text{HF}}_{(\text{gas}).}$$ (5)

Chlorapatite is ferroelectric and its phase transition occurs at a temperature estimated to be between 675 and 775 K (Rausch, 1976), and may be at ~695 ± 4 K (Hitmi et al., 1984). These temperatures match with those in Venus highlands (Seiff, 1987; Lorenz et al., 2018). Treiman et al. (2016) posit that differences in anion composition (proportions of F⁻, OH⁻, and Cl⁻), in particular the F:Cl ratio, or cation composition (substitution of Sr or rare Earth elements for Ca) in apatite can also change the Curie temperature. Apatite with a larger F:Cl ratio, as might be expected due to the reaction in Equation 5, would require higher temperatures and thus lower altitudes to display a high dielectric constant (Rausch, 1976). If the low elevated excursions observed on Maat and Ozza flows are due to apatite, this would imply that these lava flows have more fluorine (F) than chlorine (Cl).

The authors stress that the list of minerals and compounds that may exhibit ferroelectric behavior on Venus is not complete, and more work is required to measure the electrical properties of well-characterized materials under Venus temperatures at Magellan wavelengths. But of the list of compounds in Table 3, only perovskites and the mineral chlorapatite are consistent with the Curie temperatures recorded in the emissivity excursions at Atla Regio. Of the perovskites, two endmembers occur in nature: macedonite (PbTiO₃) is a rare accessory mineral associated with ore bodies (Radusinović & Markov, 1971). Lueshite (typically NaNbO₃) is a common accessory phase in carbonatites and peralkaline silicate magmas (Mitchell et al., 2014). As carbonatites have been suggested to explain exceptionally long lava flows on Venus (e.g., Kargel et al., 1994), they should be considered, although carbonatite volcanoes are rare on Earth. The other candidate perovskites are not known to occur in nature. Because of this, Treiman et al. (2016) argued that perovskite compounds are unlikely. The authors agree that because apatite is a common mineral in rocks, it is the most likely candidate ferroelectric mineral suggested to date to explain Venus ferroelectric signatures. This is also supported by the observation that the high-altitude ferroelectric emissivity excursions reported here are common on volcanoes, coronae and the tesserae on Venus (Brossier & Gilmore, 2021; Brossier, Gilmore, & Toner, 2020).

Sapas flows have an emissivity excursion, and by inference, mineralogy signature that is not seen in Maat and Ozza flows. The uniqueness of Sapas mineralogy could be explained by the fact that the volcano has its own magmatic source. Indeed, Maat and Ozza are located on Atla Regio, the highest volcanic swell on Venus. Volcanic swells are suggested to be possible hotspot sites associated to deep mantle plumes ascending from the core–mantle boundary (Stofan & Smrekar, 2005). Moreover, Atla Regio is considered to be among the most likely sites for recent tectonic and volcanic activity on Venus, which is corroborated by analysis of the gravity and altimetry data returned by Magellan (Phillips, 1994; Smrekar, 1994; Stofan et al., 1995). Sapas is a large volcano that occurs away from the volcanic swells. Isolated large volcanoes on Venus have been interpreted as secondary hotspot sites related to small-scale plumes or diapirs originating from shallower depths (Stofan & Smrekar, 2005). The authors therefore hypothesize that Sapas has distinct mineralogical properties with respect to Maat and Ozza due to different mantle source regions (Brossier, Gilmore, & Toner, 2020).

Alternatively, different atmospheric conditions above Atla Regio can account for the different emissivity patterns (temperatures, altitudes) observed there. Our current knowledge about the near-surface environment of Venus is relatively limited, and only based on a very small number of profiles from the Venera (Avduevsky et al., 1983) and Pioneer Venus probes (Seiff et al., 1980) to 12 km altitude and only one profile below 12 km from the Vega 2 lander (Linkin et al., 1986). All profiles show close agreement about the state properties of the deep atmosphere (Zasova et al., 2007), where temperature variations with latitude scatter over no more than a few K (Seiff, 1983, 1987). From theory, both latitudinal and diurnal temperature changes are expected to be very small (<1 K) in the deep atmosphere due to the high thermal inertia of the thick atmosphere (Stone, 1975). Additionally, based on Venus general circulation models, Lebonnois et al. (2018) argue that surface temperature excursions are small, supporting the idea that winds and atmospheric temperatures would not be dominant effect on weathering. Nonetheless, the authors cannot exclude the idea that local atmospheric conditions (temperature, chemistry) were different during the time of the reactions forming the minerals with high dielectric constants.

6. Emissivity as a Chronometer

If the presence (or absence) and magnitude of low emissivity excursion of surface material on Venus is a function of surface-atmosphere weathering, then the emissivity data can be used as a crude and uncalibrated dating tool, where, for a given composition and atmospheric conditions, stronger excursions indicate that minerals in the rock have had more time to weather to a mineral with a high dielectric constant in the deep atmosphere of Venus.

6.1. Temporal Evolution of the Volcanic Systems

For Sapas, the common altitude of the emissivity excursion signifies that a single mineral with high dielectric constant is present in all flows but varies in volume. In this case, the oldest flows from Sapas (S1 unit) have a higher emissivity compared to the younger flows (Figure 4a). This means that the S1 unit contains lower amounts of the ferroelectric mineral under the assumption that, if the initial volumes were similar, its stratigraphic age predicts it would have the greatest time to produce ferroelectric minerals via surface-atmosphere interactions. Alternatively, the S1 unit may have erupted with less of the high dielectric mineral or its precursor initially. With the same assumptions, the S2 unit has a higher emissivity and thus less of the ferroelectric mineral than the subsequent flow, the S3 unit. The interpretation of the youngest units, the S4 and S5 units, is complicated by their small size relative to the emissivity data footprint and the presence of fine-grained materials, however, that the lowest emissivity values for Sapas are within the S5 unit suggests that the flow has had time for the ferroelectric minerals to develop and/or has a greater initial volume of that mineral.

Maat’s M1 unit has an excursion (e4) at 6,052.9 km while the M2 unit has three other excursions (e1–3) at different altitudes from 6,053.9 to 6,056.2 km (Figure 4b). This indicates that two stratigraphically sequential flow units can possess a distinct mineralogy. In this case, the mineral responsible for the excursion (e4) exists in the initial flows of Maat but was not produced anytime later in the system. The initial flows of Maat also lack the mineral that causes the excursion (e3). The M2 unit also shows that a single flow unit can have mixed minerals; three ferroelectrics in the M2 unit record 3 Curie temperatures (Table 2). The high emissivity observed in the M3 and M4 units may be due to radar-dark materials that could be pyroclastic or ash flows that display low radar backscatter and a return to high (near-normal) emissivity (Keddie & Head, 1994), like the M5 unit, lack the ferroelectric minerals upon eruption seen in the older flow or are young enough that weathering has not produced ferroelectric minerals (Brossier, Gilmore, & Toner 2020, see Section 6).

At Ozza, the first 4 flow units (O1–O4 units) all display low emissivity excursions indicating enough time to produce the minerals with high dielectric constant. The O1 unit does not have the excursions at 6,055.7 km (e2) and 6,052.5 km (e3), indicating that it lacks the minerals that correspond to these Curie temperatures (701.5 and 730.8 K, respectively). The high T_c excursion (e3) at 6,052.5 km is only seen on the O2 unit and the flows of Ningyo Fluctus (O7 and O8 units), suggesting that these flows may share a similar source region, not represented in O1 and O6 units despite their comparable altitude ranges (Figure 5). The O5 unit has less volume of ferroelectrics, potentially because of relative youth; geologic mapping has shown that this unit may be related to possible late stage reactivation of Ozza as a shield field after caldera collapse as seen on many volcanoes on Earth (e.g., Smith & Bailey, 1968).

6.2. Ages and Sources of the Atla Volcanoes

Our findings (Figures 4 and 5) show that the ferroelectric responsible for the low elevation (below 6,053 km) excursions seen in Ozza (e3) and in Maat (e4) could have the same composition as they both occur at a similar altitude and Curie temperature (~730 K). Similarly, the high elevation (above 6,053 km) excursions seen in Ozza (e1, e2) and Maat (e1, e2) suggest a mineral with a Curie temperature of ~700 K. The common mineralogy inferred from emissivity may indicate similarity in the source regions of these volcanoes, distinct from that of Sapas. Indeed, Maat and Ozza are adjacent to each other and are both located on Atla Regio. As stated above, volcanic swells are suggested to be possible hotspot sites associated to deep mantle plumes ascending from the core-mantle boundary, whereas isolated large volcanoes such as Sapas have been interpreted to be secondary hotspot sites related to small-scale plumes or diapirs originating from shallower depths (Stofan & Smrekar, 2005). The authors therefore hypothesize that Sapas has distinct mineralogical properties with respect to Maat and Ozza due to different mantle dynamics at the origin of their formation (Brossier, Gilmore, & Toner 2020).

Atla Regio is considered to be among the most likely sites for recent tectonic and volcanic activity on Venus, which is corroborated by analysis of the gravity and altimetry data returned by Magellan (Phillips, 1994; Smrekar, 1994; Stofan et al., 1995) and thermal anomalies observed by the Venus Monitoring Camera (VMC) onboard the Venus Express spacecraft (Shalygin et al., 2012, 2015). The youthfulness of Atla Regio is further demonstrated by Basilevsky (1993) and Basilevsky and Head (2002a, 2002b) through analyses based on stratigraphic relationships between rift structures, lava flows, and crater features in the region (see Figures 8 and 9). They concluded that the late stages of tectonic and volcanic activity associated with the rift zones in Atla Regio are contemporaneous with the formation of impact craters which are young enough to preserve their associated radar dark parabolic features which are impact materials that are removed over time by physical weathering (Campbell et al., 1992; Izenberg et al., 1994). Some lava flows erupted from Ozza (O2 and O5 units) are superimposed on faults and fractures from Ganis Chasma, indicating that volcanism postdates faulting and contributes to partial infilling of the rift. However, other flows (O1 unit) are locally cut by rift structures, indicating that rifting was partly synchronous to the formation of lava flows (Senske et al., 1992; Guseva, 2016, 2019). Unlike Ozza, Maat is not disrupted by faulting, with the easternmost lava flows (M5 unit) appearing to overlie all other landforms in their paths, notably fractures from Dali Chasma (Robinson & Wood, 1993; Senske et al., 1992). This implies that these flows postdate the rifting activity. Additionally, Basilevsky (1993) and Basilevsky and Head (2002a) observed that these lava flows (M5 unit) overlap the impact crater Uvaysi on Ozza’s southwest flank (Figure 8). The radar-dark parabola associated with this crater covers both Ozza flows (O1–2 units) and Rusalka volcanic plains, making them appear darker. However, lava flows from Maat are unaffected by the parabola. Hence, these lava flows are younger than both the crater and the rift system. Basilevsky and Head (2002b) also noted that all flows from Maat (M1–5 units) are younger than surrounding craters with radar-dark halos (Figure 9). In addition to Figures 8 and 9, the reader is referred to Basilevsky and Head (2002a) (their Figures 8 and 9) and Basilevsky and Head (2002b) (their Figures 30–33) for more details about the spatial relationships between the lava flow units and the surrounding craters (ejecta, parabola and halo). If the craters with dark parabolas represent the youngest 6% of the Venus crater population (Campbell et al., 1992; Schaller & Melosh, 1998) and the average age of the venusian surface is about 300 Ma to 1 Ga (McKinnon et al., 1997; Strom et al., 1994), this means that rifting and rift-related volcanism in the region were active as recently as during the last 18–60 Ma. Recent studies advocate for a younger venusian surface of 150 Ma (Herrick & Rumpf, 2011) or 240 Ma (Le Feuvre & Wieczorek, 2011), implying a more recent activity within the last ~9 Ma. If the ferroelectric phases are formed via surface-atmosphere-interactions, this places a constraint on the kinetics of the reaction that produces the ferroelectric minerals. In this scenario the production of the ferroelectric phase via weathering at Maat would require over about 9–60 Ma (Brossier, Gilmore, & Toner 2020; Klose et al., 1992). The three latest flows from Maat (M3–5 units) have emissivity above 0.7 (Figure 4), suggesting that perhaps the last 3 flow units at Maat were produced in the last tens of Ma.

Figure 8.

Spatial relationships between Maat flows and Uvaysi crater. (a) Uvaysi crater and its radar-dark parabolic feature (parabola). The parabola darkens volcanic flows from Ozza (O1 unit) and Rusalka Planitia, whereas flows from Maat (M5 unit) appear to be unaffected by the crater deposits. (b) Uvaysi crater is superposed on Ozza flows (O1 unit) and rift-associated fractures, but its ejecta and floor are locally covered by Maat flows (M5 unit). Images (a) and (b) are adapted from Basilevsky and Head (2002a). Yellow outlines indicate crater materials (ejecta, floor, and peak all together), excluding parabola (or halo).

Figure 9.

Spatial relationships between Maat flows and surrounding craters. (a) Fossey (left) and Piscopia (right) craters and their radar-dark circular features (halos). Lava flows emanating from Maat (M1 unit) enter into the remains of Fossey’s halo without showing any darkening. (b) A protuberance of Maat flows (M1 unit) covers part of the hummocky ejecta blanket of Piscopia. (c) Melba crater and its radar-dark halo. Maat lava flows (M1 unit) embay Melba’s hummocky ejecta and cover parts of its halo without showing any darkening. Images (a) to (c) are adapted from Basilevsky and Head (2002b). Annotations are the same as in Figure 8.

The relative age of Sapas has not yet been estimated, as it is far from rifts and crater parabolas. Regarding the lowest emissivity values reported there, Sapas could be older than Maat and perhaps contemporaneous to Ozza.

7. Conclusion

Deviation of radar emissivity values from the planetary average (emissivity excursions) are seen in Ozza, Maat and Sapas montes. The emissivity excursions are spatially correlated to morphologic and stratigraphic units independently mapped using SAR data, indicating that the emissivity excursions they are related to rock composition rather than local atmospheric conditions (composition, temperature). The shape of the emissivity variations with altitude seen in the three volcanoes and Ningyo Fluctus is consistent with the presence of ferroelectric minerals in the rocks, which are highly dielectric at specific temperatures (Curie temperatures). The range of Curie temperatures are derived from the altitude and magnitude of the emissivity excursions, corresponding to values of 693–731 K over a range of elevation of 6,052.5–6,056.7 km. Low elevation excursions at Maat and Ozza indicate ferroelectric minerals at higher Curie temperatures than any excursions previously observed on Venus.

The flows of Maat and Ozza display multiple emissivity excursions corresponding to different Curie temperatures, suggesting the type or composition of ferroelectric compounds varies in each volcanic system over time. Flow events with similar emissivity excursions with altitude are not always contiguous or sequential. Sapas flows are consistent in their inferred Curie temperature through its history, unlike Maat and Ozza. If the ferroelectrics are produced by a chemical reaction between the surface and atmosphere, the presence of these minerals will correlate with age related to the (unknown) rate of the reaction. All of the flows of Ozza and Sapas have had time for this reaction to occur. The three stratigraphically youngest flows on Maat lack minerals with high dielectric constants, implying much of Maat is younger than Ozza and Sapas. The volumes of ferroelectrics in Ozza and Sapas are modeled to be greater than in Maat, also consistent with a greater degree of weathering over time. The most recent flows erupted from Maat could be younger than 9–60 Ma based on their stratigraphic relationships with crater parabolas and rift structures.

For ferroelectric compounds, the temperature and altitude of the emissivity excursion are functions of composition, while the magnitude of the excursion is a function of volume of the ferroelectric. The existence of distinct emissivity excursions at similar altitudes in the flows of Maat and Ozza montes and Ningyo Fluctus indicate that these volcanoes share common ferroelectrics, while the minerals of Sapas flows are distinct from its neighbors. Assuming the atmospheric composition is constant over Atla Regio during the eruption of these lava flows, the dielectric signatures may reflect variations in the primary mineralogy of the melts.

The emissivity patterns of the flow units of Ozza, Maat, Sapas and Ningyo indicate a minimum of 4 different ferroelectric minerals. The authors calculate minimum estimates for the volumes and types of ferroelectric inclusions responsible for that each emissivity excursion and obtain volumes in the ppm range (6–113 ppm). The Curie temperatures of derived from the altitude of the emissivity excursions are consistent with the mineral chlorapatite and perovskite oxides and inconsistent with other substances, including GeTe which has been previously proposed for Venus. Experiments on the formation, stability, and Curie temperatures of candidate ferroelectrics at Venus conditions will advance our understanding of the petrology of these volcanoes and the nature of the surface-atmosphere interactions.

Large volcanic rises like Atla Regio represent one important science target for future orbital and in-situ investigations as they are possible sites of current tectonic and volcanic activity. The proposed missions VERITAS (NASA); (Helbert et al., 2018; Hensley et al., 2020; Smrekar et al., 2020) and EnVision (ESA) (Ghail et al., 2018; Helbert et al., 2019) would return complementary, critical datasets including improved topography, SAR imaging, gravity, and infrared spectroscopy. Such data are essential to better map the radar properties, distribution, sources and stratigraphy of the lava flows at Atla Regio and place in context with improved gravity and new measurements of infrared emissivity which will constrain rock composition, thermal anomalies and atmospheric composition and variability.

Data Availability Statement

All data presented in the figures (images and shapefiles in maps, and tables of values to produce the scatterplots) are given in the online repository linked to this study (Brossier, Gilmore, Toner, & Stein, 2020). Magellan datasets used in our mapping and plotting procedures are provided in the USGS websites and described in Ford et al. (1993): Synthetic Aperture Radar imagery (https://astrogeology.usgs.gov/search/map/Venus/Magellan/Venus_Magellan_LeftLook_mosaic_global_75m), altimetry (https://astrogeology.usgs.gov/search/map/Venus/Magellan/RadarProperties/Venus_Magellan_Topography_Global_4641m_v02) and emissivity (https://astrogeology.usgs.gov/search/map/Venus/Magellan/RadarProperties/Venus_Magellan_MicrowaveEmissivity_Global_4641m).

Key Points:

• Mapped flow units in Maat, Ozza, and Sapas montes are spatially correlated to radar emissivity anomalies at multiple altitudes

• Emissivity anomalies are consistent with the presence of ferroelectric minerals (ppm scale), such as chlorapatite and some perovskite oxides

• The stratigraphy and magnitude of the emissivity anomalies support a young age (~10s Ma) for the flows of Maat Mons