VLBI astrometric and imaging observations of Active Galactic Nuclei (AGN) at K band (24 GHz) were initiated in 2002 by Lanyi et al. (2010) and Charlot et al. (2010) resulting in a K-band astrometric catalogue of 268 sources. In 2013, building from this foundation, a new collaboration was formed with the aim of enhancing the source density and sky coverage and improving the astrometric accuracy of the K-band catalogue. In 2018 the K-band astrometric catalogue was adopted as part of the third realisation of the International Celestial Reference Frame, the ICRF3 (Charlot et al., 2020). As of January 2024, the K-band Celestial Reference Frame (K-CRF) consists of 1327 relatively uniformly distributed sources - comparable to the number of regularly observed sources at S/X-band - derived from 153 observing sessions and more than 2.4 million delay observations. For sources overlapping with the standard S/X-band frame, the median precision of the K-band frame is now slightly better than the S/X-band frame. This development highlights also the growing potential of K-band geodesy. In a forward-looking approach, the K-band collaboration is charting a roadmap aimed at continual improvement in observation quality and frame accuracy (de Witt et al., 2023). Simultaneously, efforts are underway to expand the collection of K-band VLBI products, spanning across disciplines such as astronomy and geodesy, thus broadening the horizons of this collaborative endeavour.

Our work is motivated by the benefits of high-accuracy Celestial Reference Frames (CRFs) for many applications such as determining the Earthʼs orientation in space, studying the motion of tectonic plates, the alignment of radio and optical reference frames, studies of Galactic aberration, satellite tracking, orbit determination, deep-space navigation, alignment of the planetary ephemerides, and as phase reference calibrators for VLBI imaging of weak and extended sources and measurements of parallaxes and proper motions, and for its contribution to the Global Geodetic Reference Frame (GGRF), which is the subject of a recently adopted United Nations resolution on a global geodetic reference frame for sustainable development.

Why did we add a celestial reference frame at K-band?

1. To allow data from multiple independent data sets to be compared: Astrometric solutions available at different radio wavelengths facilitate the comparison and study of VLBI solutions across these frequencies. This comparative analysis helps identify and understand any discrepancies or systematic errors present in the data. Cross-frequency calibration and inter-comparison between different radio frequencies can help mitigate systematic errors,  improve the accuracy of the measurements, and provide redundancy in the data, enhancing its reliability and ensuring the robustness of the reference frame.

2. To realize the potential advantages of higher frequencies: The advantage of K-band over the historically standard S/X-band frequencies is the notable factor of 3 improvement in resolution, coupled with a generally more compact source morphology and reduced core-shift effect at higher radio frequencies. Moreover, operating at higher radio frequencies, such as K-band, facilitates observations closer to the Sun, as the solar plasma effect decreases inversely proportional to the square of the frequency. Furthermore, these higher frequencies also allow for observations closer to the Galactic Plane, as the broadening effects induced by Galactic scattering are diminished. Observing at higher frequencies should therefore permit ultimately the construction of a more accurate and more stable reference frame, which would be advantageous in many respects, in particular for aligning the VLBI frame and Gaia optical frame.

3. To meet the demands of modern VLBI systems: There is a growing need to transition towards higher radio frequencies. This shift is evident in the decline of geodetic stations equipped with S/X dual-band receivers, which is being offset by the introduction of next-generation receivers and instruments. For example, the Korean K/Q/W-band (22/43/86 GHz) receiver system, now being installed at many new sites other than Korea, is expanding the reach of high-frequency CRF work, paving the way for advanced VLBI applications. Additionally, envisaged large-scale research infrastructure projects like the next-generation Very Large Array (ngVLA) are not expected to have dual-band S/X capability. The move to higher radio frequencies is also essential to the National Aeronautics and Space Administration (NASA), the European Space Agency (ESA), and the Japan Aerospace Exploration Agency (JAXA), where deep-space missions are using Ka-band (32 GHz). Moreover, lower frequency bands, such as S-band and X-band, face increasing challenges due to Radio Frequency Interference (RFI), which significantly degrades data quality. Transitioning to higher frequency bands like K-band, which are less susceptible to RFI, provides a solution to mitigate these issues, ensuring cleaner and more reliable data for astronomical observations.

4. To serve as a crucial resource for astronomical VLBI observations at higher frequencies: Increasing the number of well-observed calibrator sources will enhance the precision and effectiveness of phase-referencing and differential astrometry for VLBI observations at K-band. Differential VLBI astrometry on water masers, vital for enhancing our understanding of the Milky Way galaxy's structure, requires well-observed calibrator sources at K-band. Additionally, many non-geodetic radio telescopes, including stations from the European VLBI Network (EVN), typically have K-band receivers. Therefore, K-band astrometric observations play a pivotal role in providing precise locations for these non-geodetic telescopes.