Imagined Temperature Sensations

Updated 9 July 2026

ITS is a heterogeneous phenomenon where temperature sensations are generated by internal imagery and cross-modal cues rather than direct heat sources.
Research using EEG, VR/MR psychophysics, and hybrid thermal interfaces shows that ITS elicits robust subjective effects and sensorimotor responses with measurable, modest offsets.
ITS bridges immersive technology, neurorehabilitation, and energy-efficient thermal comfort applications by coupling visual, auditory, and embodied cues with indirect thermal actuation.

Searching arXiv for the listed ITS-related papers and closely related work to ground the article in current literature. Imagined Temperature Sensations (ITS) designate a heterogeneous class of temperature experiences in which perceived warmth or coldness is generated, modulated, or amplified by internal imagery, multisensory context, embodiment, or indirect actuation rather than by a conventional localized thermal source alone. In the recent literature, ITS includes explicit mental temperature imagination in EEG paradigms, mixed-reality and virtual-reality thermal illusions driven by visual and auditory cues, avatar- and environment-mediated shifts in thermal comfort, and hybrid systems in which non-contact or wearable thermal interfaces shape temperature perception while remaining compatible with immersive interaction (Belichenko et al., 22 Aug 2025, Wilson et al., 2024, Weiss et al., 1 Sep 2025, Iwabuchi et al., 2024).

1. Conceptual scope and operational definitions

Across the cited work, ITS is not a single protocol but a family of operationalizations. In one line of research, ITS is defined narrowly as mental temperature imagination: participants vividly imagine hot or cold sensations on the hand for a fixed interval, in the absence of physical thermal stimulation (Belichenko et al., 22 Aug 2025). In another line, temperature sensations are altered by immersive cues in VR or MR without actual temperature changes at the contacted object, so that perceived temperature is dissociated from physical temperature (Wilson et al., 2024, Weiss et al., 1 Sep 2025). A further extension appears in AR and contactless haptics, where indirect thermal rendering or airborne ultrasound provides either partial physical support or hybrid co-stimulation, thereby situating ITS between pure imagery and direct thermotactile actuation (Iwabuchi et al., 2024, Watkins et al., 26 Mar 2025).

This plurality matters because different paradigms target different dependent variables: cortical desynchronization, subjective estimates in degrees Celsius, point of subjective equality (PSE), thermal comfort, skin temperature, embodiment, or object-matching accuracy. A coherent reading of the literature is therefore methodological rather than doctrinal: ITS refers to temperature experience as constructed by the nervous system under altered sensory contingencies.

Paradigm	Manipulation	Primary readout
Mental temperature imagination	Hot or cold imagined on the hand for 4 s, no physical stimulation	Mu-ERD over sensorimotor cortex
MR/VR thermal illusion	Visual or auditory cues with unchanged object temperature	Verbal estimate, PSE offset, comfort
Embodiment/environmental modulation	Fire/ice hands, virtual heaters, warm/cool scenes	Skin temperature, thermal perception, embodiment
Hybrid or physical rendering	Airborne ultrasound or wearable thermoelectric actuation	Temperature rise, JND, matching accuracy

The literature therefore distinguishes at least three layers of ITS: internally generated imagery, cross-modal perceptual bias, and indirect thermal rendering. This suggests that ITS is best understood as a systems-level phenomenon spanning imagery, expectation, embodiment, and thermosensory inference.

2. Visual and auditory induction of apparent temperature

Mixed-reality work has shown that classical temperature illusions can be recreated and extended in synthetic environments. In an MR study with $n=30$ , participants interacted with a real, room temperature object at $21^\circ$ C while viewing one of five overlays: Red, Blue, White, Fire effect, or Ice effect. The study reported that the color-temperature illusion can be recreated in MR and that dynamic graphics can create a new temperature-sensory illusion. Quantitatively, there was a main effect of visual effect, $F(3.35,97.17)=3.16$ , $p<0.05$ , a main effect of block, $F(1.93,55.47)=8.05$ , $p<0.005$ , and an interaction, $F(6.39,185.19)=2.4$ , $p<0.05$ . Within Block 1, Fire was perceived as hotter than Ice and Ice as colder than Blue; the blue-versus-red color effect was modest, approximately $2^\circ$ C in Block 1, and non-significant in Blocks 2 and 3. The paper interprets the color effect in terms of expectation-contrast and argues that dynamic fire/ice graphics may operate through richer semantic associations that are not fully explained by that mechanism (Wilson et al., 2024).

A later VR study quantified these effects psychophysically rather than relying only on abstract ratings. Study I used magnitude estimation with $N=20$ , and Study II used a method of constant stimuli with 2AFC and cumulative-Gaussian fitting with $21^\circ$ 0. Thermal illusions significantly influenced subjective temperature estimates in Study I, with a main effect of cue at $21^\circ$ 1. However, the directly measured PSE offsets in Study II were small: Warm VA (Steam + Sizzle) $21^\circ$ 2C, Warm A (Sizzle) $21^\circ$ 3C, Warm V (Steam) $21^\circ$ 4C, Cold A (Ice cubes sound) $21^\circ$ 5C, Cold V (Ice cubes visual) $21^\circ$ 6C, and Cold VA (Ice cubes + sound) $21^\circ$ 7C, with baseline $21^\circ$ 8C. JND values remained near $21^\circ$ 9C and were not elevated relative to baseline. The study concludes that thermal illusions robustly alter subjective judgments but only modestly affect actual perceived temperature, on the order of $F(3.35,97.17)=3.16$ 0C, and that auditory cues can be as effective as or more effective than visual cues (Weiss et al., 1 Sep 2025).

Taken together, these studies delimit an important misconception. ITS induced by purely visual or audiovisual cues can be reliable, but the measurable offset in perceived temperature can be substantially smaller than subjective scaling alone implies. The distinction between a robust subjective effect and a small psychophysical offset is central to evaluating ITS rigorously.

3. Embodiment, avatars, and virtual environments

A distinct ITS literature examines whether thermally suggestive virtual bodies and environments alter both perception and physiology. One 2025 paper proposes three controlled experiments in VR. The first tests virtual heaters, with heater On versus Off presented per hand and outcomes including continuous skin temperature, self-reported thermal perception or comfort per hand, and sense of presence. The second adapts the Rubber Hand Illusion paradigm to VR, crossing Avatar Hand Type (Fire-hands versus Ice-hands) with Stroking Synchrony (Synchronous versus Asynchronous), and measuring skin temperature, subjective thermal perception and comfort, proprioceptive drift, and embodiment questionnaires. The third uses a $F(3.35,97.17)=3.16$ 1 within-subjects design to separate Thermal Cue (Hot versus Cold scenes) from Hue/Color Temperature (4500K warm/orange versus 12500K cool/blue) while recording skin temperature and thermal perception or comfort (Kocur et al., 14 Apr 2025).

The same paper summarizes preceding studies by Kocur and Henze (2023). Those studies reported that embodying “fire hands” led to a reduction in skin temperature compared to “ice hands,” whereas the virtual environment alone did not impact skin temperature but did systematically modulate thermal perception, with participants feeling warmer in fire-themed worlds. This produces a nontrivial dissociation: physiological responses may depend more strongly on avatar embodiment, while environmental effects may act more strongly on perception and comfort (Kocur et al., 14 Apr 2025).

These results complicate any simple account of ITS as mere semantic priming. The planned use of synchronous versus asynchronous stroking explicitly tests whether thermal modulation depends on ownership rather than cue exposure alone. A plausible implication is that ITS can be strengthened when thermal meaning is attached to the represented body rather than only to the scene.

4. Non-contact and wearable thermal rendering

ITS research also intersects with interfaces that provide indirect or hybrid physical support for temperature experience. A non-contact method based on high-intensity airborne ultrasound focuses acoustic energy directly on human skin via 12 ultrasound phased-array devices (AUTD3 units), with a thermography camera (Optris OPTPI 45ILTO29T090) co-located at the center of the arrays. Stimuli were delivered to the palm at a distance of 296 mm, using either Static Focus or Amplitude-Modulated focus as a 50 Hz square wave with a duty ratio of 0.9. In preliminary experiments, static focus produced a $F(3.35,97.17)=3.16$ 2C increase after 5.0 s and up to $F(3.35,97.17)=3.16$ 3C after 30 s, while the 50 Hz amplitude-modulated condition produced $F(3.35,97.17)=3.16$ 4C after 5 s and $F(3.35,97.17)=3.16$ 5C after 30 s. Because the just-noticeable difference for thermal sensation on the palm is reported as as low as $F(3.35,97.17)=3.16$ 6C, these increases are highly perceptible. The same focal ultrasound also produces localized radiation pressure, so mechanical pressure and thermal increase are co-localized and concurrent (Iwabuchi et al., 2024).

This non-contact approach is relevant to ITS because it collapses the usual separation between tactile and thermal channels. The work explicitly argues that simultaneous presentation of mechanical and thermal stimuli can support richer haptic feedback in VR or AR and likely support the induction or modulation of imagined temperature sensation when warmth and pressure cues are aligned. Unlike Peltier plates or heat lamps, all energy transfer occurs through air, which avoids physical attachment to the skin (Iwabuchi et al., 2024).

Wearable AR systems pursue a complementary strategy. A 2025 AR device identifies three design considerations: indirect feedback, thermal passthrough, and spatiotemporal rendering. The hardware uses a $F(3.35,97.17)=3.16$ 7 grid of thermoelectric modules embedded in a flexible silicone base, each with dedicated thermistor feedback and closed-loop PID control. Human-subject experiments on 12 participants reported “All-Warm” JND $F(3.35,97.17)=3.16$ 8C (95% CI: 2.30–2.71), “All-Cool” $F(3.35,97.17)=3.16$ 9C (2.54–3.22), “Line-Warming” $p<0.05$ 0C (3.49–4.27), and “Line-Cooling” $p<0.05$ 1C (3.36–4.14). In the temperature-passthrough experiment, overall matching accuracy with the device was 84.6% and statistically significant at $p<0.05$ 2, compared with 78.9% for a bare hand, and realism averaged 5.73/7. Spatial-pattern detection was approximately 78%, and moving thermal stimuli increased enjoyment from 3.42 to 5.50 and immersion from 3.08 to 5.67, with realism of moving sensation at 5.83 (Watkins et al., 26 Mar 2025).

These systems are not pure ITS in the narrow imagery sense, yet they clarify the engineering boundary conditions under which imagined and rendered temperature can reinforce each other. The literature thus supports a continuum from fully illusory temperature to strongly physically grounded thermal rendering.

5. Neural correlates of imagined temperature

The most direct neural evidence for ITS comes from EEG. In a study of 15 healthy right-handed adults, real thermal stimulation (TS) and imagined temperature sensation (ITS) were compared using electroencephalography. TS used a palm-mounted Peltier-based tactile display to deliver Hot at $p<0.05$ 3C or Cold at $p<0.05$ 4C for 4 s, accompanied by flames or ice visuals. ITS used no physical stimulation; participants vividly imagined the corresponding hot or cold sensation on the same hand for 4 s while viewing neutral images. EEG was acquired with a 48-channel DC system at 500 Hz, and analysis focused on event-related desynchronization of the sensorimotor mu-rhythm (8–13 Hz), especially at C3 (Belichenko et al., 22 Aug 2025).

Both TS and ITS triggered significant and robust ERD in the mu-band during the entire 4 s stimulation period. The ERD was highly localized over contralateral sensorimotor cortex, notably C3, C1, C5, CP3, and CP5, with little or no significant ERD at C4 or Cz. Relative to resting baseline, ERD was statistically significant for both TS and ITS in both Hot and Cold conditions at $p<0.05$ 5. Direct ITS-versus-TS comparisons were not significant: $p<0.05$ 6 for Hot and $p<0.05$ 7 for Cold. The topographical patterns for real and imagined temperature were described as nearly identical and restricted to the sensorimotor area (Belichenko et al., 22 Aug 2025).

The principal implication is that imagined temperature sensations recruit sensorimotor cortical mechanisms in a manner comparable to actual thermal perception. The paper interprets ITS as reactivation of stored sensory representations and proposes that hot and cold imagery could expand BCI command vocabularies beyond motor and tactile imagery. For neurorehabilitation, the same result suggests a route for engaging sensorimotor circuits without overt movement (Belichenko et al., 22 Aug 2025).

6. Applications, limits, and disputed extensions

Several application domains recur across the literature. In XR, ITS is used to increase immersion, produce “temperature touch,” enrich object interaction, and modulate comfort without relying on full-scale heating or cooling hardware (Iwabuchi et al., 2024, Watkins et al., 26 Mar 2025). In sustainability-oriented VR, visual environments and avatars are investigated as tools for shifting thermal comfort and potentially reducing energy consumption in the context of global warming (Kocur et al., 14 Apr 2025). In neuroscience and human-computer interaction, ITS is proposed as a non-motor BCI strategy and as a possible component of neurorehabilitation (Belichenko et al., 22 Aug 2025).

At the same time, the literature identifies clear constraints. First, the strength of illusory ITS depends strongly on modality and evaluation method. In MR, visually induced effects diminished across successive blocks, suggesting habituation and learning that all objects were similar in temperature (Wilson et al., 2024). In VR psychophysics, subjective judgments shifted robustly, yet the actual measurable perceptual offsets were relatively small, approximately $p<0.05$ 8C, and did not always align with abstract ratings (Weiss et al., 1 Sep 2025). Second, different manipulations act on different levels: environmental cues may modulate perceived warmth without affecting skin temperature, whereas avatar embodiment may alter skin temperature in ways that do not follow naive semantic expectations, as in the reported fire-hand cooling effect (Kocur et al., 14 Apr 2025).

A more expansive and disputed extension of ITS appears in work on meditative visualization and focused attention in altered states of consciousness. That paper reports 159 attempts and 2427 operator-sensor sessions across five months, with increases of body temperature up to 38.5 C, intentional control of core temperature trends by 1.6 C, persistence lasting $p<0.05$ 9 min, and induced thermal fluctuations at the $F(1.93,55.47)=8.05$ 0 C level in external calorimetric systems with 15 ml of water for 60–90 min. It reports repeatability over 90%, including 57/63 positive targeted attempts versus 11/67 control attempts, with Chi-square $F(1.93,55.47)=8.05$ 1 and Mann-Whitney $F(1.93,55.47)=8.05$ 2. The explanatory framework offered there is a new model involving spin phenomena in biochemical and physical systems, especially spin conversion in water molecules (Kernbach et al., 2023).

Within the present body of work, this last model should be read as a specific proposed interpretation rather than as a settled account of ITS. The broader literature more securely supports three conclusions: temperature experience can be generated or biased by imagery and cross-modal context; embodiment and immersion can couple thermal meaning to both perception and physiology; and direct or indirect thermal interfaces can either substitute for or amplify ITS, depending on the target application and required effect size (Weiss et al., 1 Sep 2025, Iwabuchi et al., 2024, Watkins et al., 26 Mar 2025).