Higher surface temperatures near south polar region of the Moon measured by ChaSTE experiment on-board Chandrayaan-3

Introduction

Lunar temperatures and thermophysics dictate the stability of water-ice and volatiles^1,2,3,4,5, lunar geophysics and geology⁶, resource exploration, mission safety⁷ and aid in establishing sustainable long-term habitats on the Moon⁸. The extremely low thermal conductivity of the lunar surficial layer acts as a thermal blanket and inhibits the propagation of solar heat flux to the subsurface causing notable surface temperature variation (both spatial and temporal) and a very steep temperature gradient within the first few centimetres^9,10. While the Apollo 15 and 17 heat flow experiments^11,12,13 and LRO Diviner observations^9,14,15 have constrained this crucial surficial layer to be around 2-10 cm, an in-situ temperature profile of the lunar surficial layer is not available. Further, no in-situ measurement of temperatures is available from lunar high-latitudes and poles, since Apollo measurements are from equatorial regions only. Chandra’s Surface Thermophysical Experiment (ChaSTE) on-board Chandrayaan-3’s Vikram lander is an experiment to provide temperatures of the top 10 cm of regolith and thermal conductivity of a high latitude location in the lunar southern polar region. Similar to ChaSTE, attempts have been made earlier to deploy thermal probes, MUPUS on-board ESA’s Rosetta Philae lander for comet 67 P/CG¹⁶ and HP³ on-board NASA’s Insight mission for Mars¹⁷, by using a hammering based mole device. However, both the instruments could not be successful in penetrating to the desired depths for measurements, either due to irregular landing position (Philae) or due to surface recoil (InSight)¹⁷. While MUPUS used a hammering device, the ChaSTE needle was pushed into the soil by a rotating device, which pressed a shaft into the soil. To the best of our knowledge, ChaSTE is the first experiment to demonstrate penetration as a potential mechanism for deploying planetary thermal probes, for carrying out near surface in-situ measurements.

Experiment, observation and data analysis

Chandrayaan-3, the third Indian lunar mission had a successful soft landing at a high latitude location (69.373° S, 32.319° E), on the Moon. This location is now called as ‘Statio Shiv Shakti’ (https://planetarynames.wr.usgs.gov/Feature/16272). Chandrayaan-3 consisted of a Vikram lander and a Pragyan rover. The lander and rover carried six scientific instruments for conducting in-situ investigations at the landing site⁷, which included the elemental composition measurements¹⁸. ChaSTE is an experiment on the Vikram lander aimed at in-situ investigation of the temperatures and thermophysical properties of the top 10 cm of the lunar subsurface regolith^19,20. The ChaSTE experiment consisted of a thermal probe, an electronics module¹⁹ and a mechanism for deployment²⁰ and is detailed in the ‘Methods’ section. The thermal probe was deployed and penetrated into the lunar soil after Vikram soft landed on the Moon on 23 August 2023. After deployment and commissioning of ChaSTE observations at 12:19 UTC on 24 August 2023, the measurements were conducted continuously for the duration of the mission, till 2 September 2023 (See ‘Methods’ section for sensor configurations and phases of operation).

Results and discussion

Temperature profiles within the top 10 cm from ChaSTE

Figure 1 shows the actual in-situ surface and sub-surface temperatures for different depths within the top 10 cm of the lunar surface measured by ChaSTE at Vikram landing site during 25 August–1 September. In order to derive the actual temperatures, ChaSTE observations are required to be corrected for probe-induced perturbations as described in the ‘Methods’ section. With the effective ChaSTE probe penetration depth being around 100 ± 3 mm, the top sensor S1 is just above or on the surface and followed solar insolation with a peak temperature of ~365 K (±0.5 K) at local noon (12 h Lunar Local Time (LT)). All the uncertainties reported in this paper are maximum uncertainties expected after considering all possible errors/deviations. A house-keeping (HK) temperature sensor (mounted above S1 on the ChaSTE deployment structure) facing the Sun confirmed this observation of sensor S1 (See Fig. 2). Therefore, sensor S2 values are considered as the surface temperature. Measurements from sensors S2, S4, S6–S10 were taken to derive the temperature profile and for all further analysis and interpretation, as the orientations of sensors S1, S3 and S5 are different (See the ‘Methods’ section). The peak surface temperature of 355 K ( ± 0.5 K) is higher than Diviner observation of 330 K ( ± 3 K) for the landing site. A temperature difference of nearly 85 K, 80 K and 55 K was observed at 9:42 LT, 12:00 LT and 15:00 LT, respectively, between the sensor S2 and the bottom most sensor S10 in the regolith. It is important to note that ChaSTE measurement is a point observation, while other estimations are regional based on either energy balance at that latitude^2,21 or are derived from coarse resolution observations (~240 m per pixel, Diviner⁹). ChaSTE provides the advantage of capturing local scale temperature variations continuously over a fraction of a lunar day (during 9:42 LT–15:00 LT). To understand the diurnal variations in surface temperature, we carried out simulations using our established PRL 3D thermophysical model^10,22 by fine tuning the model for the Vikram landing geometry, local topography, actual configuration and observational conditions of the ChaSTE experiment. The diurnal surface temperature variations from model simulations were compared with ChaSTE measurements and Diviner observations for understanding the local scale thermal behaviour at the Chandrayaan-3 landing site.

**Fig. 1: ChaSTE measured temperature profiles using ten sensors at ten different depths, at the Shiv Shakti point on the Moon from ChaSTE passive experiment during 25 August to 1 September.**

**Fig. 2: Comparison of ChaSTE Sensor S1(red) and S2 (green) observations with a nearby HK data (blue) from Vikram lander (sensor placed just above S1 on the ChaSTE deployment structure).**

Local topography, slope and landing geometry

Images from Orbiter High Resolution Camera (OHRC) on Chandrayaan-2 orbiter²³, before and after landing and the location of ChaSTE are given in Fig. 3. From the post-landing images taken by the NAVCAM camera onboard Pragyan rover (See Fig. 4a), it was confirmed that the lander was positioned looking towards a small crater/depression with ChaSTE probe oriented in North-West direction. The angle between the orientation of ChaSTE and true North direction, known as the north aspect angle, was estimated to be ~10°, as depicted in Fig. 3. At this orientation, with the lander residing on a small shallow crater rim, ChaSTE probe was found to be penetrated on a tilted surface, which has a sunward local slope angle of ~6°. The local slope angle was obtained from lander inclinometer data which is also independently verified with slope derived from OHRC DEM data. This was confirmed from analysing various other NAVCAM images as shown in Supplementary Fig. 1 and OHRC Digital Elevation Model (DEM), confirming that Vikram has landed on a crater rim with ChaSTE oriented towards a local sunward slope. The topography of the landing site for the model is generated from OHRC DEM, which has the highest spatial resolution of 0.25 m per pix²³. Image showing Vikram lander with ChaSTE deployed and the reconstructed landing perspective for modelling are shown in Fig. 4a, b. Figure 4c represents a vertically exaggerated 3D surface of ChaSTE location generated from the Chandrayaan-2 OHRC DEM along with a schematic representation of various geometrical parameters used to calculate the incident solar flux. For modelling at local scale, we selected an area of 4 m × 5 m with Vikram lander location at the centre as shown in Fig. 4b. Reconstructed south looking perspective of the lander used for model calculations is shown in Supplementary Fig. 2. Since the illumination conditions at high-latitudes and poles change drastically due to the local topography, the actual corresponding solar flux with respect to the local slope angle is calculated (See Supplementary Methods) and provided as one of the inputs to the model.

**Fig. 3: High resolution images from OHRC before and after Vikram’s landing on the Moon.**

**Fig. 4: NAVCAM captured image of Vikram lander and its reconstructed perspectives.**

Diurnal variability and comparison with ChaSTE observations

The surface temperature at the ChaSTE landing site obtained from the model compared with ChaSTE in-situ measurements is shown in Fig. 5. A good agreement between the ChaSTE observations (solid red) with model (solid blue) confirms that the unusually high temperature reported by ChaSTE is due to its lander’s orientation and probe insertion on a terrain with a local slope of ~6°. The model derived diurnal temperature showed a peak surface temperature of 355 K ± 5 K at local noon with night-time temperature converging to 105 ± 5 K.

**Fig. 5: A comprehensive plot showing in-situ observations and model derived diurnal temperatures at ChaSTE penetration point.**

In order to decipher the surface temperature at a flat terrain (0° local slope) i.e. the surface temperature that ChaSTE would have observed if it would not have landed on a slope, we used in-situ temperature measured by a thermistor (housekeeping sensor) mounted outside the lander’s bottom deck and looking at the lunar surface which is continuously illuminated by solar radiation during the observations. This sensor radiatively receives heat from the lunar surface, just below the lander, similar to the observations made during Apollo and Surveyor missions^12,24,25,26. We found this peak temperature to be ~320 K ± 1 K. Model simulations for 0° slope condition is also plotted in Fig. 5. The modelled peak temperature at noon for 0° slope is ~332 K ± 5 K, which is consistent with in-situ observations (from housekeeping sensor beneath the lander). To the best of our knowledge, ChaSTE has not only provided the first in-situ observations of temperatures at a high latitude region of the Moon, but also demonstrated that the topography and slopes at smaller scales of few metres can result in notable temperature differences. These effects become prominent as we move towards poles, an important aspect to be considered for future exploration.

For comparison and to substantiate our findings, we have identified Diviner radiometer footprints obtained at the nearest spatial distances to ChaSTE measurement location. Methodology similar to the one reported in an earlier study⁷ is used for this study. Since Diviner has a resolution of ~240 m per pixel, finding the nearest point could only make a generic and more realistic way of comparison. For this purpose, absolute temperatures of nearest locations to ChaSTE within 500 m × 500 m area were derived from Diviner brightness temperatures provided in Supplementary Data 1 (using an emissivity of 0.86 of WAC albedo map²⁷) and average values for a given local time are plotted in Fig. 5 (Yellow Squares). From Diviner observations, a peak temperature of ~330 K ± 3 K was found at a location nearest to ChaSTE which is consistent with our model derived value for a flat (0°) terrain near ChaSTE. Our model-derived day and night temperatures for 0° slope are also in a very good agreement with Diviner observations as shown in Fig. 5. This further ascertains that local slopes at this fine scale influence the surface temperatures which are otherwise not captured by Diviner. With no in-situ observations available, surface temperatures calculated by earlier thermal models provided a wide range of temperatures viz. ~290–300 K²⁸, 268–275 K⁹, 300–320 K² at ~70° latitude, which also suggests that the local scale thermal environment on the Moon could be highly different, further supporting ChaSTE observations.

Implications on sub-surface water-ice migration and stability

It is established from earlier studies that the local thermophysical environment is an important factor controlling the presence and stability of water-ice on the Moon^{2,3,29,30,31,32,33}. With ChaSTE offering a unique opportunity for in-situ investigation of the local thermophysical environment, we investigated the plausible conditions for surface/subsurface stability of water-ice at the landing site. Modelling studies showed that surface temperatures ranging from 110 K to 114 K are suitable conditions for cold-trapping of water ice, particularly at polar regions^34,35,36. It is also suggested that water-ice can migrate to the subsurface due to strong thermal pumping, if the diurnal maximum temperature (T_max) is above 120 K and the mean temperature (T_min + (T_max-T_min/π)) is below 105 K³⁶. Also, the presence of a dry regolith layer can reduce sublimation rates, potentially burying water ice deposits (due to any transport mechanisms) under a few centimeters thick layer^2,37,38,39. Such buried water-ice can remain stay put as long as the temperatures are maintained below cold-trapping limits, especially during lunar nights. ChaSTE observations of much higher surface temperatures imply that water ice on the surface is unstable. But, locations in the vicinity with different slopes and sun-orientation can have different temperatures and thermophysical environments allowing for the stability of water-ice. To explore this aspect using the ChaSTE in-situ measurements, we have carried out model simulations at the landing site for varying slopes, sun-orientation and illumination conditions as shown in Table 1. It is evident that the surface temperature can vary dramatically with changes in local topographical orientation with poleward slopes receiving much less solar flux and thus exhibit lower surface temperatures as shown in Supplementary Fig. 3. Such temperature conditions could be suitable to harbour water molecules, especially at the subsurface. Model derived diurnal temperatures for poleward local slopes of 10^o (dark red), 14° (pink) and 18° (dark green) are shown in Fig. 5, which also shows the temperature zones for sub-surface water-ice migration (green band), it’s stability and cold-trapping temperature limit, within the model uncertainties. While sunward slopes would not allow water-ice migration, larger poleward slopes (>14°) allow for water-ice migration to the sub-surface as shown in Fig. 5. Further, it is possible to cold-trap these water-ice deposits, as the night time temperatures are well-below the cold-trapping temperatures of water-ice. Therefore, it is possible that certain regions around ChaSTE location having larger poleward slopes might have water-ice deposits near surface (within top 20 cm) or at the subsurface (below 1 m), and even possibly as micro cold traps²⁹. Thus, ChaSTE observations have shown that at high-latitudes, sunward slopes are considerably warmer and poleward slopes just about a metre apart could be much cooler, providing an environment conducive for the presence of water-ice within the shallow sub-surface. Therefore, unlike poles, high-latitude regions with larger poleward slopes can also be potential sites for water-ice prospecting for future exploration and in-situ resource utilisation. Such sites would be technically less challenging for exploration in comparison to extreme polar sites, but scientifically equally interesting for enhancing our understanding about the water ice quantification, distribution and migration.

Table 1 Model calculated peak solar flux, maximum (max) and minimum (min) temperatures for various local slope orientations, diurnal peak temperatures and implications for stability of water-ice³⁷

Full size table

Methods

ChaSTE thermal probe and deployment mechanism

ChaSTE thermal probe is designed using a special composite material of a very low thermal conductivity and sufficient mechanical strength, in order to withstand various launch loads and wide range of operating temperatures on the lunar surface^19,20. The sensor used for temperature measurement is a custom-designed Platinum RTD (Resistance Temperature Detector) sensor with a wide measurement range and linearity. Ten platinum RTD (PT1000) sensors (S1–S10) are mounted on the probe at various distances along the axial direction to measure the temperature within the top 10 cm of the lunar surface. A thermofoil heater is also wrapped around the probe at the position of the ninth sensor (S9) for thermal conductivity measurements. Figure 6 shows the structural diagram of the ChaSTE probe with mounting positions of the 10 sensors and the heater. Sensors S1, S3 and S5 are mounted on -Pitch side, while the other sensors S2, S4, S6–S10 are mounted on +Yaw side with respect to the Vikram lander spacecraft. While the total length of the ChaSTE probe is 400 mm, the penetrating part of the ChaSTE probe is 140 mm, with an effective distance from sensor S1 to sensor S10 being 94 mm (centre to centre). The remaining length of the probe above the S1 sensor is extended in length with the same diameter and lobes, to ensure the structural integrity of the probe as shown in Fig. 6. Harness from all 10 RTDs and heater mounted on the probe are routed through the inner hollow section of the probe. The front of the probe has a conical tip, made of Titanium alloy, to assist in penetration into the lunar surface. The other end of the probe is connected to a penetration mechanism through a vespel adaptor. The probe and mechanism are further housed in a Glass Fibre Reinforced Plastic (GFRP) housing with a flap at its bottom to protect the probe as well as the mechanism from lunar dust. The penetration mechanism is a ball screw-based mechanism which converts the rotary motion of the motor into a translation motion of the probe through helical threads on the ball screw shaft.

**Fig. 6: Structural diagram of ChaSTE probe with mounting positions of the 10 RTD Sensors and Heater.**

ChaSTE electronics module

The electronics of ChaSTE contains Front-End electronics (FE), DC-DC converter, Motor Drivers and Processing electronics. The FE is responsible for acquiring the raw sensor data from the 10 RTD sensors and heater, signal conditioning, heater motor current sensing and digitisation of data, after which it is provided to processing electronics (PE) for further handling^19,20. ChaSTE electronics containing box is kept at -Pitch side under a thermally controlled environment inside the lander. The PE controls the operations of probe deployment and its penetration into the lunar soil and interfaces the payload data with the lander for telemetry and tele-command.

ChaSTE deployment and observation timeline

Chandrayaan-3’s Vikram lander had a successful touch down at 12:34 UTC on 23 August 2023 at Shiv Shakti point. After accomplishing initial operations of commissioning of the lander mission, including the Pragyan rover roll down, the deployment and commissioning of ChaSTE was carried out at 12:19 UTC on 24 August. ChaSTE payload penetration and science operations were initiated at around 20:30 UTC. ChaSTE instrument operations are primarily accomplished in four phases—Deployment, Probe Penetration, Passive (Temperature observation) and Active (Heater) experiments, as briefly described here. The penetration operation was carried out in two phases. The first penetration lasted for 45 min, ended at 20:07 UTC on 24 August inserting the probe to a depth of 90 ± 3 mm. The second penetration of about 2 min ended at 15:33 UTC on 25 August, making the effective penetration depth of the probe to be 100 ± 3 mm. The complete timeline of operations of ChaSTE experiment is given in the Table 2. The measurements were considered for passive experiment after second penetration. Measurement of temperatures from the 10 sensors continued till the active heating experiment started (See Table 2).

Table 2 Timeline of ChaSTE operations and key diurnal phases on the Moon

Full size table

ChaSTE deployment

ChaSTE Probe along with its mechanical assembly was kept in stowed configuration, anchored to the +Y Panel of the lander using a frangi bolt, during the flight, till landing. Once the ChaSTE instrument operations were initiated, the first operation was to release the deployment mechanism, by firing the frangi bolt at 20:45 UTC on 23 August. On 24 August, the payload was initially switched on after which the deployment sequence was initiated through ground command at around 12:21 UTC. The probe was deployed slowly in steps till it achieved the desired operational configuration, normal to the lander and the position was locked through a latch switch. Thereafter, the ChaSTE instrument was kept on and data was continuously recorded. As a part of system integrity and health check, temperature stability of all the sensors was monitored before penetration and was found to be within ±0.1 K.

Thermal probe penetration

After deployment, probe penetration was initiated to move the probe slowly in forward direction towards the lunar surface till the tip of the probe touched the lunar surface. The temperature variation of the bottom most sensor (S10) located near the probe tip is continuously monitored and used to identify the probe’s contact with the lunar surface. Once probe contact was confirmed, the probe was further moved down in steps till all the temperature sensors of the probe were inside the lunar surface. The distance moved by the probe for penetration into the regolith is inferred from the motor counts using hall sensors present inside the brushless DC motor used in the penetration mechanism. The Hall sensors measure the rotation of the motor shaft and the motor shaft rotations are converted to the linear distance using the gear reduction ratio and the lead on the ball screw shaft.

Passive experiment

Passive experiment started after probe penetration was completed and continued till the start of active experiment. During this phase of the experiment, temperatures from all the 10 RTD sensors on the probe were continuously measured. These observations are used to obtain the variation in temperature within the top 10 cm of the lunar surface, as a function of both depth and time.

Active experiment

In this mode of the experiment, the heater placed at Sensor 9 location is provided with a known amount of power and the propagation of the heat from the heater is monitored to derive the bulk thermal conductivity of the soil. The details and results of this active experiment are discussed separately in another paper.

Observation of probe-induced perturbations on ChaSTE measurements

Instantaneous temperature recorded by the bottom most sensor (S10), at a depth of the lunar soil, as the probe penetrates gives the actual temperature at that depth for that instant of time. After complete penetration, the depth profile of temperatures measured by S10 sensor during its traverse is expected to be similar to the one measured by all sensors of the probe. This is because, the time taken for penetration operation was just 45 min wherein no notable change in temperatures is expected due to solar forcing. However, a clear deviation is observed in the temperature profiles measured by S10 sensor (during penetration) and the one measured by all the sensors S1–S10 (static) as shown in Fig. 7. After first penetration operation instantaneous temperatures measured by the bottom most sensor S10 at different depths as it penetrates down is shown in Fig. 7. This figure also shows temperature profiles observed by all the sensors (S1–S10) at various instances of first and second penetration operations. The blue curve represents the temperature profile measured on 24 August, 19:20–20:07 UTC by the tip sensor (Sensor 10) of ChaSTE probe as it penetrates into the lunar regolith. This curve represents the actual depth profile of temperatures of the regolith at that point of time. The pink curve shows the depth profile of temperatures recorded by all the sensors (S1–S10) immediately after first penetration. Comparison of blue and pink profiles clearly indicate a probe induced effect, particularly by the probe tip, in the measured temperatures as the probe penetrates down. This is because of the differences in thermophysical properties of the probe material, its tip and the surrounding regolith. However, the temperatures seem to retain an equilibrium condition soon as evident from the temperature profiles an hour after first penetration (green) and 4 h before second penetration (dark red). The solid red curve shows the temperature profile measured after second penetration on 25 August, 16:34 UTC by all the 10 sensors. Comparison of profiles shown in Fig. 7, clearly show the probe induced effect that needs to be corrected to derive actual temperature of the regolith. This effect of the probe on the measured temperatures is also validated using laboratory experiments and further estimated using a three-dimensional finite element thermophysical model^7,10 as discussed below. The observation is corrected for the probe effect to derive the actual temperatures.

**Fig. 7: Temperature profile as observed during first and second penetration operations.**

Experimental verification of probe-induced perturbations on temperature measurements

In order to verify the effect of probe material on the measured temperatures, we conducted a series of experiments using our Lunar Simulation Chamber Facility at PRL^40,41. These experiments were carried out under a well-controlled simulated lunar environment using ChaSTE probe and electronics, identical to that flown onboard Chandrayaan-3 Vikram lander. A pre-baked sample strata of 200 mm depth and 160 mm diameter using a lunar highland simulant soil, Sittampundi Anorthosite^10,42, of heterogeneous grain sizes ranging 40–250 microns was used for carrying out the experiment (Supplementary Fig. 4). The probe was placed inside the simulant soil in the same configuration as encountered by ChaSTE experiment on the Moon. To mimic solar insolation conditions and the incident flux, a DC powered halogen lamp with adjustable intensity and look angle, was used as a solar simulator. Series of 10 additional reference temperature sensors at depths identical to that of ChaSTE Sensors were also placed in the simulant soil to monitor the unperturbed profile. Proper care is taken in placement depths and positions of the reference sensors to ensure that they are measuring temperatures at the same depths and placed well-apart to avoid any mutual disturbance and edge effects. After verification of electrical connections, the chamber was subjected to slow pumping to achieve simulated lunar conditions following an established experimental protocol^40,41. Once the required experimental conditions were achieved, solar insolation is simulated by switching on the halogen lamp and temperature from all the 10 sensors of the probe and the reference sensors were continuously measured for about 2.5 h till the effect of probe is evident in the temperatures monitored by all the sensors. As it was impractical to experimentally recreate the time-varying (diurnal) incident solar radiation on an infinite lunar regolith surface in the lab, these experiments were carried out for shorter durations with an aim to only verify the effect of probe material on the measured temperatures. Solar insolation at a fixed inclination angle on a finite regolith medium is simulated in each of these experiments. The results from one of these experiments is shown in Supplementary Fig. 5. It can be clearly seen that the temperatures recorded near the probe are lower than those observed without the probe at the same depth indicating a perturbed temperature field around the probe due to the presence of the probe in the soil. Lower temperatures recorded near the probe are attributed to the relatively high thermal inertia of the probe material with respect to the surrounding regolith as already reported from modelling studies earlier^43,44,45,46. The experiment was repeated several times to ascertain this phenomenon and also for different solar insolation conditions. A similar trend is observed in all the experiments thus establishing our inference.

Correction of probe-induced perturbations using a 3D thermophysical modelling

To establish the experimental observation of probe-induced perturbations as described above and correct it, we used our Lunar 3D thermophysical model developed at PRL^7,10. Temperature deviation factors derived from the model simulations were used to correct the perturbed temperatures and to derive the actual temperatures measured by ChaSTE experiment on the Moon. Model simulations were carried out considering an identical deployed configuration and operating conditions of ChaSTE experiment on the Moon. Details of model simulations are given below.

Model geometry

We modelled every element of the ChaSTE probe, including smallest part of the assemblies, such as sensors, bonding material as shown in Supplementary Fig. 6. This was done to ensure that the collective effect of all the materials is accounted. To simulate the surface regolith conditions of the Moon, local scale topography of the landing location from high resolution DEM from Chandrayaan-2 OHRC data²³ is used. The lunar surface was considered as a multi-layered structure, with different thermophysical properties, with an outermost porous layer. The probe and the sensors are assembled into a cavity generated inside the 3D lunar surface with a perfect thermal contact between the probe and the surface thus mimicking the final deployed configuration of the ChaSTE probe. The assembled geometry used for the model is shown in Supplementary Fig. 7. This final deployed state of the probe inside the regolith is considered for all simulations and was set up following an approach described in an earlier study¹⁰.

Initial and boundary conditions

We have developed equations¹⁰ to simulate the movement of the Sun during the daytime as seen at ChaSTE location on the Moon. These equations were used along with the ambient radiation boundary conditions to simulate the thermal environment experienced by the probe at the landing site. The probe assembly modelled inside the regolith is oriented in such a way that the sensors S1, S3, S5 are facing to EAST where the sun shines first during the early morning and the sensors S2, S4, S6–S10 are towards North and face the sun during mid-day (See Supplementary Fig. 7). The material properties of all the modelled geometries considered in the simulation are given in Supplementary Table 1.

Model results and derivation of correction factors

The solar insolation reaching the top surface of the regolith as a function of the local time and the corresponding latitude is defined in the model. The surface and sub-surface are defined as porous media with appropriate material properties as defined in Supplementary Table 2. The heat is conducted and radiated across the porous medium from top to bottom layer of the regolith. The probe and the lunar surface also radiatively exchange heat with the ambient. The pressure inside the pores of the regolith geometry is set to a negligible value to mimic ultra-high vacuum condition on the Moon and to remove any convective effects that the model may consider. The model was developed using commercially available COMSOL^® Multiphysics Software with heat transfer module (Version 6.2). Model simulations were carried out for more than 200 Earth days (~7 lunar days) till convergence is achieved and the temperatures obtained from the last simulation day were used for analysis.

Diurnal temperatures for all the sensors S1-S10 were obtained from simulations, that are expected to be perturbed by the presence of the probe. To estimate the undisturbed temperatures at the same sensor locations, simulations were carried out considering only the lunar surface (without the probe) with the same configuration, initial and boundary conditions. Diurnal temperatures for all the sensors were obtained in this case as well, that represent the true (unperturbed) lunar temperatures. The ratios of the temperatures obtained for each sensor in these two cases gives the perturbation factors for each measurement depth (S1–S10). As the perturbation is induced by the presence of probe inside the regolith, the perturbation factor is primarily a function of the probe material and independent of the regolith properties, which was also verified through several simulations. In-situ measurements from ChaSTE are then multiplied by these temperature deviation factors to obtain the unperturbed or actual temperatures.

Diurnal profiles of perturbed and unperturbed temperatures obtained from simulations, for four sensor locations, two at top, S1 & S2, and two at bottom, S9 and S10, are shown in Supplementary Fig. 8. The probe affects the in-situ temperature profile due to its presence inside the lunar surface by increasing the effective thermal inertia of the surrounding medium thereby smoothing the diurnal temperature variation. The top few centimetres of regolith layer has a very low thermal inertia as compared to that of the probe and thus the measured temperature profiles show a large deviation from the probable in-situ temperatures. However, at larger depths, where the regolith thermal inertia is higher, the probe effects are negligible. The temperature deviation factor due to probe perturbation is the ratio of the expected temperatures (only regolith) and the measured temperatures (regolith with probe deployed) as mentioned earlier. The probe-induced perturbations in the temperature field within the soil are depicted in the Fig. 8. Figure 8 also shows the comparison of the expected temperatures (black) and the corrected temperatures (within brackets in blue) using the derived coefficients thus validating the approach. The derived correction coefficients are provided in Supplementary Data 2.

**Fig. 8: Model derived perturbation in the temperature field around the ChaSTE probe.**

3D Thermophysical modelling of temperatures at ChaSTE location

A 3D thermophysical model¹⁰ utilising the actual topography and incoming solar flux was used to derive the diurnal surface temperatures at ChaSTE location. The procedure for calculating solar flux at the ChaSTE location is given in the Supplementary Methods. The model considers the lunar surface as a multi-layered structure with a porous upper layer. The region is taken to be an area of dimensions 4 m × 5 m × 2.5 m and the thermophysical parameters, such as the thermal conductivity, bulk density is given as temperature dependent functions (see Supplementary Table 2). As heat is exchanged between the surface and the surrounding space including the lander, the ambient temperature becomes an important boundary condition for a three-dimensional modelling, particularly for a surface with topography and local slopes. During daytime, each facet of the topography will have a different view angle with respect to the Sun resulting in multiple reflections and re-radiation in the ambient environment as a result of pronounced directionality and non-diffuse nature of the lunar surface⁴³. This was also evident from in-situ thermal engineering data and cable thermocouples observations during Surveyor and Apollo Missions^12,24,25,26. A temperature value of ~250 K near to dawn from energy balance (when the surface starts responding to the diurnal solar insolation) and also as observed by Chandrayaan-3 in-situ engineering data (housekeeping data from a temperature sensor placed outside of the anti-sun panel of the lander, away from the lunar surface and looking at lunar ambience) is considered as initial ambient temperature condition. As the ambient temperature drops due to cooling of the surroundings during night, the ambient temperature boundary condition for the model is represented as a step function of 250 K during the day and 4 K at night for the ChaSTE location. While modelling for other locations, value of dawn temperature for a given local slope, estimated from energy balance and solar flux modelling, is used as a true peak ambient temperature in this step function. The solar flux and ambient temperature functions used in the model for ChaSTE location are plotted in Supplementary Fig. 9. The entire geometry is meshed into extremely fine surface elements with size varying from 0.1 to 10 cm. Each model simulations were run for ~2000 Earth days with a time step of 0.1 Earth days to achieve convergence. The model was developed using commercially available COMSOL^® Multiphysics Software with heat transfer module (Version 6.2). Since the local scale modelling is computationally intensive, all the simulations were carried out using one Petaflop Param Vikram–1000 High performance cluster computing facility at Physical Research Laboratory, Ahmedabad.

Conclusions

ChaSTE experiment onboard Vikram lander of Chandrayaan-3 has provided the first in-situ temperature profiles near south polar region of the Moon, as per the best of our knowledge. ChaSTE measured lunar regolith temperatures up to a depth of 10 cm at a high latitude highland region of the lunar south pole over a major fraction of a lunar day. The peak surface temperature at Shiv Shakti landing site is 355 K, a temperature relatively higher than predicted by Diviner observations, which is attributed to ChaSTE penetration on a Sun-ward (North ward) facing slope region. Temperatures estimated using a thermophysical model, fine-tuned to the Chandrayaan-3 local geometry, show a good agreement with in-situ measurements. Lunar surface temperature measured on a flat surface, about a metre away from ChaSTE location, is 332 K, which is consistent with orbiter based remote sensing observation (330 K). ChaSTE observations suggest that the lunar surface temperatures show a significant spatial variability at metre scales. The sunward slopes are found to be recording much higher temperatures (~30 K or more) in comparison to flat locations or poleward (anti-sunward) slope regions. A relationship between the local slope and expected surface peak temperature has been derived. The interesting outcome is that the high latitude sites with local slope greater than 14° in poleward direction might offer similar environment as polar sites for accumulating water ice at shallow depths. The ChaSTE measurements provide a lot of new information about the temperature distribution within the outermost 10 cm regolith layer of lunar high-latitude highlands. ChaSTE findings not only indicate fine scale spatial variability in regolith temperatures but also suggest that high-latitude regions are potential sites for scouting water-ice, resource prospecting and habitation. Such sites are not only scientifically interesting but also pose less technical challenges for exploration in comparison with regions closer to the poles of the Moon.