NASA launched the CloudSat and CALIPSO satellites together on the same rocket on April 28, 2006. CloudSat carries a Cloud Profiling Radar (CPR) which allows it to pierce dense cloud systems to better understand the content and structure of clouds and analyze the convective processes within them. CALIPSO carries a lidar instrument which allows it to observe thinner cirrus clouds and measure aerosols in the atmosphere. Joint observations by the two missions allow scientists to study and better understand the relationship between clouds and the tiny particles in the atmosphere, called aerosols.

CloudSat and CALIPSO were designed to fly together, in formation, taking coincident measurements of clouds and aerosols. Soon after launch, they took up residence in the Afternoon Constellation (A-Train), where they flew in formation with each other as well as other member missions in the A-Train Constellation. The formation flying concept of the A-Train missions provided nearly simultaneous observations of the earth system below as they circled the globe, crossing the equator at 1:30pm local time every day.

CloudSat is managed for NASA by the Jet Propulsion Laboratory, in Pasadena, CA. The CPR was provided by JPL and the Canadian Space Agency and the spacecraft was provided by Ball Aerospace, in Boulder, CO. CALIPSO is operated by the Centre National d’études Spatiales (CNES), in Toulouse, France and is managed for NASA by Langley Research Center, in Hampton Virginia, which also provided the Lidar instrument. The spacecraft was provided by CNES. These international and geographically distributed teams have worked tirelessly together over the years, coordinating with each other during nominal operations as well as through several spacecraft anomalies.

CloudSat and CALIPSO flew together in the A-train for almost 12 years. During the summer of 2017, CloudSat suffered an issue with one of the reaction wheels. That wheel was taken offline and the mission was redesigned to operate on three wheels. In order not to endanger the other A-Train satellites, a decision was made to exit the A-Train Constellation. In February 2018, CloudSat exited the A-Train, maneuvering to an orbit safely below the A-Train and leaving its sister satellite CALIPSO behind. The future of the CloudSat / CALIPSO partnership became a topic of much discussion in the following months. The question: was it more beneficial to the science community for CALIPSO to remain within the A-Train, or should it exit the A-Train and form a new constellation with CloudSat? While coordinated science with the A-Train provided many benefits, CALIPSO’s strongest scientific synergy was with CloudSat, as is evident from more than 2000 scientific papers written about the composite science data gathered by the two satellites.

Soon the voices of the joint CloudSat and CALIPSO science community became clear. They wanted the partnership to continue. There was no debate and the decision was unanimous. CALIPSO would lower its orbit too and resume formation flying with CloudSat at a newly dubbed C-Train (CloudSat and CALIPSO) orbit.

This paper will discuss the process by which CloudSat and CALIPSO reshaped their partnership. Re-establishing formation required both satellites to conduct a series of orbit lowering and inclination adjust maneuvers to re-align the satellite ground tracks. It was a cooperative effort, involving international collaboration amongst the mission partners. A delicate dance was orchestrated and executed by the teams over the course of several months in the Fall of 2018. By November 2018, both satellites had completed the necessary maneuvers and as of this writing the C-Train is a reality. While there were a few bumps along the way, the spirit of international cooperation and collaboration made the concept of the C-Train a reality. CloudSat and CALIPSO are together again. It is hoped that the sister satellites will continue to operate in the new C-Train for years to come and that other missions can learn from the CloudSat / CALIPSO experience.

Area C - Board 280

Now a day more a more missions benefit to fly in tandem with another satellite. The scientific performance depends on the geometry between both satellites, usually being more beneficial to fly as close as possible one from each other, however, a minimum distance is required in order to guarantee a safe formation flying.
Having an initial trade-off on the safety and complexity of the tandem with respect to the resulting performance is essential already in the early stages of the mission.
The purpose of this paper is to give some basic guidelines on tandem design. It is noted that being just guidelines not all the cases can be comprised and the definitive control shall be based on deeper analysis. Nonetheless, they are appropriate to perform a quick and methodical first analysis.

In general, the strategies for tandem maintenance can be grouped in three main categories,
- Uncoordinated control box: It is the simplest control. It is based in defining boundaries for the MLST and the ground-track for each satellite with respect to their reference orbits. As long as these rules are respected the required geometry between both satellites is maintained without any kind of coordination between the missions.
- Coordinated control box: If a tighter control is required a minimum coordination is needed. The most simple and efficient is to coordinate the out of planes maneuvers, as they not very frequent and easily predicted. The in-plane maneuvers remain uncoordinated and each satellite shall keep its reference ground-track within a certain dead-band.
- Master-Slave control: It requires a full coordination between both control satellites, as the slave needs to follow the master dynamics. This strategy is though when a very tight control, within a few seconds, is required.

On top of the nominal control, the safety in case of an anomaly occurring on one or both satellites shall be also addressed. A passive safe formation is always preferred, that is, that in case of anomaly the tendency of both satellites is to drift apart. Unfortunately, a passive safe formation is not always possible and under certain anomalies and/or strategies, intervention from ground is needed in order to avoid a conjunction. To derive safety requirements the parameter to identify is the minimum reaction time required to avoid such a conjunction. It shall be guaranteed that from any initial configuration when the anomaly occurs this reaction time is respected.

In addition, the complexity of the operations also need to be taken into account, especially when both satellites are controlled from different mission control centers.

In conclusion, the presented guidelines are meant to help the scientific community to have an initial estimation on the feasibility and complexity of the proposed tandem between two satellites.

Area C - Board 281

ABSTRACT

The concept of “trains” of satellites with instruments on board different satellites observing the same area or atmosphere at rather short time interval to exploit synergies is well established. The most well known example is the NASA A-train. In a train configuration, satellites have the same orbital period, and a similar orbital plane, ensuring observations from the same area at a quasi-constant time (along track) difference.
The concept of Tandem missions or having tandem mission phases are also widely known and implemented. In a tandem mission, the satellites have again the same orbital period, the same earth-fixed ground track and a quasi-constant observation time difference. Early examples of tandem missions were ERS-1 and ERS-2, with a one day observation time difference, ERS-2 and Envisat with a 30 minutes observation time difference. Recent examples are Jason-3, flying in tandem with Jason-2 with along track observation time difference of several minutes, down to the recent Sentinel-3 tandem phase with 30 seconds observation time difference.
So the techniques to establish a quasi same-time observation of the same area is well understood but require the satellites to have the same orbital period, and a rather similar orbital plane. The question is whether it is possible to achieve synergetic observations with
An alternative orbit configuration.
In case it is not possible to place a satellite in the orbital plane of the identified tandem satellite providing the synergetic observations, what are the alternatives? The same question can be posed for the orbital height. The answer depends on the requirements for the synergistic observations.
In case the observation area is expected to be global, or the synergistic observations are required continuously, the classic concept of tandem mission is the most suitable. However, in case the areas of interest for the observations are limited to a particular region with a more relaxed observation time difference requirement or in case the common observations are needed at a rather low frequency (daily and longer), then other concepts than tandem missions can be considered.
In order to illustrate possible alternative orbit configurations for satellites operating in synergy, this paper will present two different example solutions:, one covering the concept of satellites flying on different orbit planes and another showing the case when the satellites are at different LEO altitudes.
The first case, analyses the synergism between a LEO satellite in dawn-dusk 18:00 LTAN orbit and MetOp-SG in a 21:30 h LTAN orbit.
To meet the stringent thermal requirements for the CIMR candidate mission, the CIMR design would strongly benefit from a dawn-dusk orbit, which is not subject to eclipses during many months in the year. CIMR however also would benefit from synergistic data observed by MetOp-SG at its nominal Mean Local Solar Time of ascending node at 21:30.. By limiting the synergistic area to the polar regions and accepting time differences of 10 minutes, together with the rather wide instrument swath of both instruments, a non tandem solution is possible.
The solution identified for CMR will be presented in detail in a dedicated paper.
An example of synergism between a LEO satellite flying at difference altitudes originates from the instrument interference experienced between Sentinel-1(A/B) and Radarsat-2, and further studied for Sentinel-1 and the Radarsat Continuation Mission. The difference in orbit altitude introduces variations in the relative speed between the satellites, leading to changes in the geographical location where one satellite ‘overtakes’ the other (beating pattern).
The geographical location and repetition frequency between these collocation opportunities depends on the resonance between the repeat cycles of both satellites. The larger the altitude difference the more often (and shorter in duration) the common observations will occur.
If the synergy requirements are limited to an occasional cross-verification between measurements of the same area taken at the same time, the concept of a beating pattern can be used to ensure that regularly the instruments on the two satellites will observe the same area at the same time.

Area C - Board 282

An Excel based software tool for cost estimation of future SAR satellites was developed. Graphic outputs include estimated cost based on specified parameters of the mission. The initial release includes cost changes due to mission delay, and cost versus different confidence levels. The current version of the tool supports large, moderate, small and mini (below 250kg) SAR satellites and their constellations. Different types of antenna and imaging modes, such as polarimetric, interferometric, high resolution wide swath (HRWS), ground motion target indicator (GMTI), 3D radar, can be selected for the mission. The user interface of the software, called the Multi Aperture SAR satellite COst estimation Tool (MASCOT), includes help tips to guide decision makers into selecting appropriate parameters or future satellite mission designs. Early designers can analyze trade-offs between complexity, cost and user requirements. Cost changes can be analysed based on adding advanced performances (such as GMTI or HRWS) to missions at a fraction of the overall mission cost.
The technical and operational parameters (cost drivers) of the SAR mission to meet requirements were identified. A work breakdown structure covering all phases of a program was defined and ground rules and assumptions helped understand and compare cost against known space programs. Parametric cost estimates are derived based on statistical relationships between historical costs and other program variables such as system physical or performance characteristics, contractor output measures, or personnel loading. Parametric cost estimate was based on a simple linear regression model.
A very important parameter considered in this tool is the satellite development cost which can be up 2-3 the unit cost and up to 6 times for high-tech programs. Cost for launching varies depending on the orbit, satellite size and mass, and the type of rocket (launching vehicle). The development heritage varies for different missions and it is defined as the percentage of a subsystem that is identical to one or more previous spacecraft. If the cost change is not significant then an input from design experts may be required to assign proper heritage value. An important variable is the cost per kg for a satellite hardware on the ground. This cost was estimated using known relationships and development heritage with the STD about 20%.
The cost of mission consisting of multiple satellites (e.g. RCM, SAR-Lupe, CSK) includes additional expenses due to the simultaneous operation of several satellites (e.g. mission guidance system). The number of satellites is also taken into consideration. So when number > 1, the cost will increase due to higher production, launching and operation costs. The usage of cost models with mentioned input parameters were used to generate estimates for identifying the most cost effective design. Cost and schedule risk analysis is performed to capture uncertainties.

Area C - Board 283