About this Article

Svea Hernandez (sveash[at)stsci.edu) and the COS team

Published April 29, 2026

Ultraviolet (UV) spectroscopy is a powerful and essential tool for characterizing the physical properties of solar system objects, stars and exoplanetary systems, high-energy transients, the interstellar, circumgalactic and intergalactic media, supermassive black holes, and a wide range of other astrophysical targets. The Cosmic Origins Spectrograph (COS) aboard the Hubble Space Telescope (HST) is one of only two far-ultraviolet (FUV) spectrographs currently in operation, providing the community with critical medium- and low-resolution UV spectroscopic capabilities. Moreover, it is the most sensitive UV spectrograph ever flown.

The COS FUV microchannel plate detector was originally designed for a nominal five-year operational lifetime when it was installed in 2009. However, thanks to the sustained efforts of engineers and scientists at STScI and NASA, COS has far exceeded this expectation and will remain operational through the 2030s.

The primary life-limiting factor for the FUV detector is gain sag, the natural degradation of detector efficiency due to cumulative photon exposure. This effect is mitigated by increasing the detector’s high voltage and by shifting target spectra to new, more pristine locations on the detector, called lifetime positions (LPs). Each move to a new LP requires a mini-commissioning effort to optimize positioning, enable observations, and ensure accurate calibration. In 2017, the COS team adopted an operational strategy called COS2025 (COS ISR 2018-16) that extended the FUV channel’s projected lifetime by optimizing how detector real estate is consumed; isolating regions of heavy airglow exposure; and triggering LP moves based on the slower-sagging continuum regions.

A set of new strategies were developed to extend the operational lifetime of the COS FUV channel to 2030 and beyond. First, the instrument began operating at multiple lifetime positions per cycle to further optimize the life of each LP. Second, the use of new areas of the detector was enabled by new split-wavecals in HST Cycle 29 (2021; see COS ISR 2023-15), at the cost of slightly increased overheads. In fall 2025 at the start of Cycle 33, the team enabled two new LPs using these two strategies: LP7, hosting the most widely used setting; G130M/1291, along with G130M/1055, 1096, and 1222; and LP10 hosting all G160M configurations. The low-resolution G140L grating remains at LP3, while G130M configurations with central wavelengths at or redder than 1,300 Å remain at LP5 (see figure). With the completion of the calibration phase of LP7 and LP10, the COS team has now implemented a total of eight lifetime positions over the nearly 17-year operational history of the instrument.

Central to future lifetime extension efforts is a newly developed and adopted dynamic patching scheme of the spacecraft flight software, which removes a previous flight software limitation of eight LPs. The implementation of this framework, known as LP-infinity, allows the team to enable and commission as many LPs as resources and available detector real estate permit. It also enables new LPs without requiring flight software updates to the instrument, reducing the cost of commissioning each new LP.

The FUV detector real estate is primarily limited by the detector size, and each mode is also constrained by the aperture mechanism. With each newly enabled LP, the available FUV detector real estate is reduced, limiting the number of future LPs. Despite this challenge, the COS team continues to develop approaches to exploit remaining pristine detector regions. In summer 2025, the team initiated the first stages of the next lifetime-extension milestone by exploring potential regions to host low-resolution G140L observations at a future LP11 without additional overheads from split-wavecals, starting in Cycle 34 (November 2026). In parallel, the feasibility of relocating the heavily used G130M/1291 setting to a new region near the detector center (LP12) was investigated and is planned for Cycle 35. Such a move would restore higher spectral resolution and lower overheads relative to LP7 for this critical FUV mode.

This potential advancement for G130M/1291 is enabled by a newly developed gain-sag flagging methodology in the COS data processing pipeline (CalCOS). With the previous approach, detector pixels were conservatively flagged when they reached a modal gain of 3 or below, corresponding to a flux loss of > 5%. Even with the two-zone extraction method used by CalCOS (COS ISR 2015-03), this conservative threshold resulted in many pixels along a given column being rejected even when the integrated flux loss across the spectral profile at that position was negligible. The refined methodology flags pixels only when they contribute to an integrated count loss exceeding the maximum achievable signal-to-noise ratio for a given LP.

In practice, this means each LP can be used longer and more of the detector area can be used for science observations, directly expanding the available real estate for future LPs. The combination of this refined flagging approach with ongoing efforts to identify and enable LP11, LP12, and future LPs promises continued operations of the world's premier FUV spectrograph well into the 2030s.

The continued availability of high-sensitivity ultraviolet spectroscopy with Hubble is key to addressing many science priorities from the Astro2020 decadal survey. Beyond the adjustments described here, additional operational approaches are being explored to ensure high-sensitivity medium-resolution FUV spectroscopy remains available through the next decade, supporting both Hubble’s ongoing science program and the future science with NASA’s next flagship telescope, the Habitable Worlds Observatory.