There are 7 basic instrument operational modes that can be commanded with the SWI application software:

OFF: All instrument subsystems including the DPU are switched off. Consequently, there is no housekeeping data and no telemetry generated by the instrument itself. The instrument is in this mode during launch and Cruise Phase, except during calibration campaigns (e.g., NECP, PCWs, Earth and Moon flybys).

The allowed transitions between these instrument modes are illustrated Fig. .

During the Science Phase (Jupiter tour and Ganymede orbits), Juice will operate on the basis of 16 h of science operations per day, followed by 8 h for communications with the Earth ground stations (downlink and uplink activities). Because of telemetry budget limitations, ESA has defined the number of 100 telecommands (TC) as the limit for uplink per science day per instrument. This limit makes it impossible to operate SWI with individual sub-unit commands, because a single science observation lasting just a few minutes would already require tens to hundreds of those commands. An observation indeed requires several steps to be taken, like the tuning of the receivers, setting the pointing mechanism, pointing to the science target, performing the required ON-OFF integrations (including pointing to the cold sky for OFF integrations), and making the relevant hot-cold calibration integrations (including the relevant FLM movements). Even the simplest observation, i.e., a nadir stare without any repetition of the integration loop, requires a minimum of 16 commands. Early in the development phase of the instrument, it was therefore decided by the instrument team to adopt the concept of observation modes, similarly to Herschel-HIFI . These are a set of scripts, pre-loaded in the DPU memory, that can be invoked by single TC, once the instrument is set in the SCIENCE Mode. Each script is a suite of sub-unit commands, which can be grouped in repetition loops, in order to produce a complete observation (e.g., 2D map). The details of the operation of each single script execution are then uniquely governed by the list of parameters supplied by the TC.

SWI currently has a total of 34 observation modes implemented in its DPU. Observation modes can be grouped in two categories: calibration modes meant for characterization and monitoring of the instrument and its behavior, and science observation modes meant for science measurements. A mode/script is fully defined by a set of parameters which concerns, for example, tuning, receiver and spectrometer setup, mechanism movements for pointing. All modes are provided with details in the Appendices  and .

The details on how to execute a given science observation have been defined partly based on theory and partly on experience from other microwave or infrared instruments in space (e.g., ). It is necessary first to verify if the scripts are functional on the ground and then to test inflight if these observation modes provide science data with the expected quality. This mode validation thus constitutes part of the Cruise Phase operations of SWI.

SWI is an instrument aiming to observe surfaces and atmospheres in the Jovian system to measure composition, temperature and wind speeds. To achieve all these goals, the instrument was designed with all the flexibility (and complexity) as that found in ground-based sub-mm observatories, including, for instance, pointing pattern, spectral tuning, coverage and sampling, integration times, integration repetition factors, etc. All the calibration and science observation modes described in Sect.  can be fed with the appropriate parameters to match a given measurement goal. Each observation is thus unique and must not only use one of the modes with the appropriate list of parameters, but it must also carry a series of metadata that provide the necessary information to enable meaningful data reduction and analysis. The description of the concept of operations of SWI is given in the next section.

The planning of operations during any of the Cruise Phase opportunities, including the LEGA, follows several steps. In a first step, the Mission Operation Center (MOC) defines the boundary conditions of the operations: date/time, duration, allowable data volume, etc. Instrument teams are then asked for a preliminary Activity Plan (APL) with operation names and descriptions, estimates of number of TC, duration and data volume, and applicable constraints. MOC usually proposes a preliminary breakdown of the data volume envelopes per instrument. In a second step, instrument teams submit an initial sequence of operations, in the form of a preliminary list of Payload Operation Requests (POR). At this stage, there can be trade-offs between instruments, managed by MOC. Once the final envelopes are agreed upon, the teams are requested to submit their final PORs several weeks before execution of the operations.

During the Science Phase, iterations with instrument teams are managed by Science Operation Center (SOC, ). SOC initially provides a segmentation of the trajectory, which prioritizes operations on a science-driven basis. SOC then handles the various instrument team deliveries, starting with the Observation Plan (OPL). The OPL is a list of instrument observation requests, which are backed up by a science plan. Instrument teams can request observations under two statuses: prime or rider. The prime status implies the instrument has constraints on the spacecraft pointing, while the rider status bears no such implication. After harmonization of the OPLs from all the instrument teams, a skeleton spacecraft Pointing Timeline Request (PTR) is generated and SOC designates which instrument team has to design the spacecraft pointing for each time block of the timeline. Once the final PTR is produced, SOC provides a skeleton Instrument Timeline (ITL), which provides the information on spacecraft resources (power, data rate and data volume). Some operations can then still be removed to comply with, for example, the total power available to instruments at a given point on the timeline.

Because the LEGA operations were managed by MOC, and because there was no spacecraft pointing allowed during the whole LEGA timeline, the setting of the whole timeline of operations did not necessitate to go through the various steps that will be required in the Science Phase. Some trade-offs had to be found for the closest approaches to the Moon and the Earth, but, as will be discussed in the next sections, SWI was one of the few remote sensing instrument capable of performing observations of the Earth and the Moon in the days that followed the closest approach to the Earth, by using its pointing mechanisms.

The planning of SWI observations requires accurate pointing information that relies on the interpretation of the mission SPICE (Spacecraft, Planet, Instrument, C-matrix, Events) kernels (https://www.cosmos.esa.int/web/spice/spice-for-Juice, last access: 23 April 2026), which are regularly updated and delivered by ESA. This is true for all modes except the calibration modes.

Geometrical calculations derived from the SPICE kernels must be performed individually for each script in order to obtain accurate pointing information for the observations to be scheduled. Other target-specific geometric information that cannot be derived a posteriori from the commanding parameters such as, for instance, target name, target size, sub-spacecraft latitude and longitude, and local time, must also be included in the metadata of an observation.

In practice, the metadata of an observation are not uplinked to the spacecraft. Only the list of TCs containing the sequence of observation modes with their relevant list of parameters is sent for operation execution. For a given observation, those metadata, the selected observation mode and the parameter list are kept in a database of planned observations. They are used upon the downlink of the telemetry binary files (both housekeeping and science data) by the SWI telemetry-to-raw (TM2RAW) pipeline to automatically create the proper file structure and to populate it with the raw data. A high-level description of the SWI TM2RAW pipeline is shown in Fig. . This is absolutely necessary, because, for example, the binary data file from a 60 min 2D map with 10 × 5 on-source points with 5 s integration time per point will have a different structure and contain a different amount of data (e.g., number of spectra) compared to a 10 min nadir stare. A second implication is that each calibration and science observation mode must have its dedicated data reduction pipeline. It is planned for the near future that the SWI telescope pointing information, although transcribed in the science metadata will also be retrievable from dedicated instrument attitude SPICE kernels.

Consequently, the SWI Science Team has opted for a unique observation identification number for each planned observation, similarly to what was made for Herschel observations (e.g., ; ; ). These so-called ObsIDs (observation identification numbers) serve this purpose of tailoring an observation, defined by its observation mode and list of parameters, with its mandatory metadata from the planning stage down to the data reduction stage. They uniquely identify, characterize and naturally encapsulate the SWI data/observations. We should also highlight here that SWI data will be calibrated and archived in chunks identified by the ObsIDs.

The preparation of the planning and uplink file deliveries by the SWI Team follow slightly different process for ongoing Cruise Phase operations and the future Science Phase after the JOI. This distinction is driven by external procedural approaches of MOC and SOC that are required during the different phases. Nevertheless, for SWI there is a strong overlap, also in the software used to produce the necessary commanding files and meta-data. In general, SWI planning relies on the selection of observations based on science priorities, and on the subsequent computation of the instrument parameters for each of the selected observation and corresponding observation mode. This is achieved with a suite of observation planning tools, which are described in the following sections. They constitute the SWI Commanding pipeline. The full flowchart summarizing the various steps from the initial inputs to the submission of PORs is shown in Fig. .

The Cruise Phase of Juice began right after its successful launch on 14 April 2023. It will last for about 8 years, until its arrival at Jupiter in July 2031. The evolution of angular sizes of the Sun, Venus, the Earth, Mars, and Jupiter, as seen from Juice during the Cruise Phase, are shown in Fig. . The first three months were used to commission the spacecraft and its instruments, during the NECP. The subsequent long interplanetary phase leading to Jupiter Orbital Insertion (JOI) comprises several regular Payload Checkout Windows, meant for periodic functional checks of instruments, performance verification and calibration. It also involves four planetary flybys to send Juice to Jupiter . These gravity assists enable controlled deviations in the spacecraft's trajectory, ultimately placing it into a heliocentric orbit tangential to Jupiter's orbit by July 2031, ensuring a successful JOI.

From the planning and operational point of view the Earth gravity assists manoeuvres provide great analogues for future Galilean moon flybys, testing SWI's relevant operational timelines. The first of these encounters, the LEGA, occurred in August 2024 and was actually a double gravity assist, with a flyby of the Moon on 19 August 2024, with a closest approach altitude of about 700 km, followed 24 h later by a flyby of the Earth, with a closest approach altitude of about 6800 km. Further payload operations were possible in the three days that followed the closes approach to the Earth. In total, the LEGA segment resulted in the execution of a total of 166 observations by SWI, for 225 observations planned. Some observations were not executed following an instrument safe mode event that is explained in Sect. .

The general strategy for the combined flybys was three-fold: (i) to validate in representative close flyby operating conditions the use of several key observation modes, (ii) to perform spectral, and spatial calibration observations using SWI's own pointing utilizing known strong lines and/or continuum, and (iii) search for new lines in the Moon exosphere and in the Earth atmosphere. It should be noted that it was the first time the atmosphere of the Earth was observed from space at frequencies around 1200 GHz. A general overview of the Moon and Earth observations is provided in the next sections.

The key characteristics of the Moon flyby in the context of SWI pointing are shown in Figs.  and . The spacecraft attitude commanding around ± 50 min the closest approach were restricted to keep optimal conditions for the geophysics (in-situ) instruments. The +Z axis of Juice was fixed essentially in a nadir view during this period. Because the lunar distance changed very rapidly near closest approach, it was not feasible to perform extensive mapping of the Moon's emission across different latitudes and longitudes or to evaluate the SWI limb-staring, scanning, or mapping modes. Therefore, we opted for an unconventional use of the SWI_2D_MAP_OTF_V1 around closest approach, by manually setting the instrument antenna in its nadir direction (aligned with spacecraft +Z axis), commanding only a single map row, and setting mechanism offsets to zero in-between subsequent raw points so as to keep the antenna pointing unchanged. Furthermore, implementing the shortest possible integration time for the CTS, i.e., 1.5 s, and defining a large “map” size has effectively enabled us to record a full track of measurements while the Moon drifted below SWI and the spacecraft during the flyby. The use of the OTF mapping mode for this observation (ObsID 227) also bore the advantage of avoiding antenna movements for the cold sky position in-between two subsequent integrations.

Once the Moon eventually moved out of the field-of-view (FOV) of the instrument, a second execution of this observation was scheduled. ObsID 228 required repointing the SWI antenna by 54° back (against the direction of the spacecraft velocity vector) using the AT mechanism. By doing so, SWI kept recording Moon spectra for approximately 45 additional minutes. For both ObsIDs (227 and 228), we tuned the receivers to the frequencies of the H₂O lines at 557 and 1113 GHz to test those tunings, because they are crucial for the Jupiter and Galilean moon observations during the Science Phase.

The Moon flyby observations are summarized in Table  of Appendix .

Similarly to the Moon flyby, the Juice remote sensing instrument platform (the +Z platform) maintained a near-nadir pointing orientation for several minutes around its closest approach to Earth. During this time SWI performed four observations (ObsIDs 229-232, see Table  in Appendix ), two on either side of the closest approach. The first and fourth were similar to those executed during the Moon flyby, i.e., two strips of SWI_2D_MAP_OTF_V1 mode manual setup. Key characteristics and the timing of the four SWI observations are shown in Fig. .

The first observation used nadir pointing of the antenna, while the fourth observation (post-closest approach) required to repoint the antenna towards the Earth. In both cases, the receivers were tuned to the frequencies of the H₂O lines at 557 and 1113 GHz to further test those tunings with a target atmosphere that contains H₂O. It should be noted that many other lines were also potentially detectable in the spectral windows covered by those two tunings.

The second observation, just before the closest approach, benefited from the largest relative size of the Earth atmospheric limb. Because this maximized the potential for detection of weak lines, the SWI team designed a limb scan using SWI_MOON_LIMB_SCAN_PS_V1, with the receivers tuned to an HDO line. This observation also enabled the validation of the use of this mode in a close flyby situation. The third observation, just after the closest approach, was designed with the SWI_MOON_LIMB_STARE_PS_V1 to observe another couple of weak lines from H217O and H218O, and also enabled the validation of using this mode in close flyby operating conditions. For both observations, we chose to compute the relevant pointing parameters (tangent altitude for the limb stare, altitude range and steps for the limb scan) for the mid-observation time, because of the rapidly changing size of the Earth and its limb. Because of these size changes and the inertial pointing of the spacecraft, these observations obviously suffer from pointing drift that will have to be carefully taken care of at the data analysis stage.

A few hours after the Earth closest approach, a rotation of the spacecraft of 180° around its +X axis enabled SWI to have the Earth and the Moon within the reach of its mechanisms until the end of the LEGA operational phase, i.e., three days after the Earth closest approach. A comprehensive set of calibration and science observations had been prepared to be run over the three days. The observation strategy is detailed hereafter and comprises generic, Moon and Earth observations.