WFI Performance

This section of the WFI Handbook describes the current understanding of the science performance of the WFI, with a focus on detector-level effects that impact calibrated imaging and spectroscopic data products.


The WFI focal plane is populated by eighteen H4RG-10 detectors, representing the most advanced IR detector technology available for the Mission. Meeting Roman’s science requirements requires calibration of these detectors to unprecedented precision, supported by an extensive ground-based characterization program.

Rather than presenting isolated articles for each individual detector effect, the WFI Performance section organizes effects by their scientific impact and discusses them together, reflecting their physical and calibration interdependencies. This structure is intended to help users understand how multiple detector effects combine in real WFI data.

This section emphasizes qualitative understanding and scientific context. Users seeking authoritative, versioned, and quantitative performance metrics should consult the Roman Space Telescope Technical Information Repository v1.2 (July 2025), which provides summary statistics for key performance parameters.

The WFI Performance documentation is under active development. Content may evolve as ground testing and commissioning continue. Users are encouraged to reference specific versions when citing or adopting performance information.

Performance information from ground testing of the Focal Plane Array is available in the Roman Technical Repository and in the WFI Quick Reference article. The technical repository also includes the serial-ID mappings between the Sensor Control Unit (SCU) and the Sensor Chip Assembly (SCA), which are particularly useful for users working with ground test data.



Summary of WFI Performance Content

The table below summarizes current and planned content in the WFI Performance section, organized by scientific impact.

Table of WFI Performance Content 

Current RDox ArticleCurrent Article ScopeFuture Additions and non-RDox References
WFI Performance (This article)
  • Overview of Current and Planned Content for the WFI Performance Section
  • Primer on H4RG-10 detectors

Pixel Behavior Families
  • Pixels with Lower Signal Accumulation
  • Pixels with Higher Signal Accumulation 
  • Other Anomalous Pixels

Instrumental Noise
  • Dark Current
  • Read Noise
  • 1/f Noise
Sources of Pixel to Pixel Variation
  • Linearity and Saturation
  • FlatFields
Non-Ideal Detector Effects
  • Burn-In
  • Persistence
  • Interpixel Capacitance (IPC)
  • Charge Diffusion
  • Brighter-Fatter Effect (BFE)
  • Count Rate-NonLineatity (CRNL)

Image Artifacts
  • Coming soon



Detector Technology Background

A comprehensive description of H4RG detector technology in the context of detector theory is provided by Mosby et al. (2020). The sections below highlight selected performance-relevant impacts associated with unique fabrication and design features of the H4RG-10 detectors used in the WFI.

The observational programs undertaken to characterize and calibrate these effects are described in the WFI Characterization Activities section of this Handbook. Detailed discussions of the algorithms applied to WFI data are provided in the Data Handbook, particularly in articles describing the Exposure Level Pipeline.


Overview of H4RG-10 Technology 

The Wide Field Instrument focal plane is populated with 18 H4RG-10 detectors. The H4RG-10 detectors are HgCdTe 4096 pixel by 4096 pixel photodiode arrays with a 10 micron pixel pitch. Prototype detectors were extensively characterized at the Detector Characterization Laboratory (DCL) at Goddard Space Flight Center under space-relevant environmental conditions (thermal vacuum, vibration, acoustic, and radiation testing). Detailed results are presented in Mosby et al. (2020) and Mosby et al. (2025).

Roman detector development was a decade-long program with key early investments, dozens of detectors fabricated and screened, then characterized and tested at GSFC through the Detector Technology Advancement Program (DTAP). The testing undertaken by the Roman project advanced the H4RG-10’s technology readiness level (TRL) to TRL-6 (Mosby et al. 2020). The Table of Teledyne Infrared Detectors provides context on the evolution of these detectors into the Roman era (adapted from Schlieder 2022, presentation).  

The remainder of this RDox article highlights some of the difference in the detector technology as compared to previous generations used in HST and JWST. Many of these physical characteristics correspond to differences in the detector performance or calibration procedures relative to the prior technology generations summarized in Table of Teledyne Infrared Detectors. More detailed discussions of these differences, and their relationship to meeting Roman requirements, are given in Mosby et al. (2020), Mosby et al. (2024), Mosby et al. (2025).


Table of Teledyne Infrared Detectors 

Detector GenerationH1RH2RGH4RG
Mission/Instrument UseHST/WFC3-IRJWST/NIRCam
JWST/NIRISS
Euclid/NIST
Roman/WFI
Development2000-20072002-20142011-2021
Pixel Array Size1024 by 1024 2048 by 20484096 by 4096
Physical Pixel Size18 microns18 microns10 microns
Quantum Efficiency 

~90%

1.0 to 1.7 microns

~90%
at 2 microns
~90%
0.8 to 2.1 microns
Dark Current

< 0.05 e-/s/pix

< 0.01 e-/s/pix< 0.005 e-/s/pix
Noise

12 e- RMS (16 reads)

6 e- RMS (1000 s)5-6 e- RMS (180 s)
Graphic


H4RG Architecture 

Infrared detectors are fundamentally different from silicon CCD detectors because it is not possible to have both photon detection and electronic readout on the same chip. The photon-sensitive mercury cadmium telluride (HgCdTe) layer is bonded to a silicon Read-out Integrated Circuit (ROIC) and the combination is referred to as a "hybrid". The hybrid is connected to a mechanical mount with electrical connections and the combination is referred to as a "Sensor Chip Assembly" or SCA. Throughout the WFI Handbook, the terms SCA and detector are used interchangeably. 

The precise mixture of HgCdTe, specifically the fraction of cadmium, can be varied to engineer a specific bandgap energy. For Roman's desired cutoff wavelength of approximately 2.5 microns, the fraction of cadmium was tuned to 0.445 (Mosby et al. 2020). Other characteristics in the HgCdTe layer induce an electric drift field within the layer that improves quantum efficiency (QE) and count rate non linearity (CRNL). 


Anti-Reflection Coating 

An anti-reflection coating acts as an interference filter and ensures that in-band light is transmitted with high efficiency. This is the last optical surface in the WFI. The actual details of the anti-reflection coating are proprietary, but the impact of the coating on light transmission is not. The transmission as a function of wavelength is shown in the Figure of the Anti-reflection Coating Transmission, which is reproduced from Figure 3 of Mosby et al. (2020). Panel A shows the transmission, as a percentage of incident light, as a function of wavelength from roughly 400 nanometers to 2600 nanometers. Three angles of incidence are shown: 0 degrees, and 21 degrees with differing polarizations designated as “S” and “P”. S-polarized is transverse electric, with the electric field perpendicular to the plane of incidence and P-polarized is transverse magnetic with the electric field in the plane of incidence. Overall, the variation in the transmission by the angle of incidence is small. Panel B shows the impact of differences in the anti-reflection coating thickness; these differences occur in the fabrication process, and a difference range of 20% of the optimal thickness is shown (e.g., 10% more and 10% less than optimal). These impacts produce significant changes in the throughput, particularly at blue wavelengths. Differences in anti-reflection coating thickness will be present across each detector and across the ensemble of the 18 detectors in Roman's focal plane array.  The result of this variation in anti-reflection coating thickness is variation in the effective blue edge at the pixel level. Significant effort in WFI Ground Testing Campaigns characterized the effective blue edge of the WFI Optical Elements and their impact on science return.

Figure of the Anti-reflection Coating Transmission

Transmission curves, as a percentage of incident light, as a function of wavelength for the anti-reflection coating on the H4RG-10. Panel A shows three transmission curves for differing angles of incidence with perpendicular to the surface, or 0 degrees) in blue, and 21 degrees S and P polarized in dotted red and dashed red. The approximate bandpass edges for the WFI Optical Elements are shown for comparison. The overall variance in transmission due to the angle of incidence is small. Panel B shows three transmission curves that show the impact on transmission for typical variations in the thickness of the anti-reflection coating caused by the deposition process (factors of 0.9, 1.0 and 1.1 are shown in blue, dotted red, and dashed red, respectively). The impact of the thickness of the anti-reflection coating does induce significant transmission differences, in particular in defining the blue edges of the bandpasses. Significant effort in Ground Performance Testing characterized these edges and their impact on science return. This figure is reproduced from Figure 3 of Mosby et al. (2020)


Surface Passivation Recipe 

The H4RG includes a novel surface passivation recipe from Teledyne (the details are proprietary; Mosby et al. 2020). The recipe, known as PV3, was necessary because Roman is passively cooled and operates at warmer temperatures than, for example, JWST and Euclid, but aims to achieve low-noise detector performance (Mosby et al. 2025). One impact of using PV3 is to reduce the impacts of persistence at these higher operating temperatures (roughly 20 times reduced from WFC3/IR and 10 times reduced from NIRCam), but the change in formula may also have played a role in temperature-dependent changes to dark performance observed in ground testing (Mosby et al. 2024).




For additional questions not answered in this article, please contact the Roman Help Desk.







Latest Update

 

Updates for the reorganization of the Instrument Handbook
Publication

 

Initial publication of the article.