INFN wiki

Researcher: Dr. Mustafa M. Rafiei

Project Title: Raw AMS Data Analysis for Space Radiobiology and Dosimetry

Contracting Agency: ASI

Grant Duration: 16 December 2024 – 15 December 2025

Last Updated: — Alessandro Bartoloni 2025/06/05

Summary of Activities:

Time allocated to general & INFN administrative procedures, IT setup, institutional onboarding, and safety courses. Preliminary definition of research activities initiated highlighitng SEP and in particular proton , Helium and Heavy Nuclei space radtion components associated to particular events happened in 2021

🧪 Summary of Activities: January was dedicated to acquiring essential technical, scientific, and experimental knowledge to support the research project. Activities were structured into technical IT training, theoretical background studies, and system-level training related to the AMS-02 experiment.

💻 Technical Training: Completed training modules in C++ programming and RStudio for data processing and statistical analysis.

Familiarized with software tools used in cosmic ray research and experimental data handling.

📚 Scientific Background & Literature Review:

• Studied foundational topics in cosmic ray physics, emphasizing Solar Energetic Particles (SEPs) and their significance in space weather and astrophysics.
• Conducted a literature review on SEP detection methods and historical observational data.

🛰 AMS-02 Experiment Training:

• Participated in video sessions covering a general overview of the AMS-02 experiment, its scientific goals, and operational principles.
• Gained insight into the housekeeping data and science data streams from the International Space Station (ISS) to the experiment databases.
• Reviewed data transfer workflows, from raw telemetry to organized scientific datasets.

🧩 Sub-detector Systems Overview:

Studied the design and functions of the main subdetectors onboard AMS-02:

• TRD (Transition Radiation Detector)
• ECAL (Electromagnetic Calorimeter)
• Tracker
• TOF (Time-of-Flight System)
• RICH (Ring Imaging Cherenkov Detector)
• Electronics and DAQ (Data Acquisition) subsystems

📑 Research Planning:

Definition of the primary objective would be identifying and characterizing particle species relevant to space radiobiology, especially in the context of short-term solar energetic particle (SEP) events.

The partciles of this identification includes:

• Protons
• Alpha particles
• Iron and/or other heavy nuclei

These particles are particularly interesting due to their role in low-energy (<10 GV) space radiation environments, where biological impact is highest during transient solar events.

🧪 Summary of Activities: February marked a critical phase in enabling analysis capability through formal training and infrastructure deployment.

🎓 NAIA Onboarding and Training Event: Participated in the NAIA onboarding event, organized by V. Formato at the INFN TOV Division.

The event provided structured training on the NAIA framework used for AMS-02 raw data analysis, including:

• NAIA architecture and module structure
• Data flow, reconstruction logic, and integration points
• Code and version control best practices

🛠 Technical Setup and Configuration: Successfully installed and configured the NAIA environment on an AlmaLinux 9-based system.

Deployed the full NAIA stack on the CNAF computing platform, including:

• Secure access setup
• Data pipeline configuration for AMS-02 raw data
• Compatibility checks and test jobs

🔍 Early Testing and Validation: Conducted test runs of NAIA modules on both AlmaLinux 9 and CNAF platforms.

Confirmed access to AMS raw data samples and validated the integrity of the pipeline.

📑 Research Progress: Integrated NAIA capabilities into early-stage planning of the research activities, with a focus on processing SEP-related proton flux data.

Began outlining the analysis chain for selected time intervals of interest.

🧾 Technical Guide Preparation and Documentation Finalized and internally released the first version (v1.0) of the "Technical Guide for Extracting and Processing AMS-02 Data Using ROOT, NAIA, and HTCondor".

This guide, co-authored by A. Bartoloni and M. M. Rafiei at INFN Roma, provides a structured reference for NAIA-based AMS-02 data analysis workflows.

Covered:

• CNAF and CERN access setup
• Remote access procedures to ui-ams virtual machine
• Best practices for SSH key generation and secure tunneling
• AlmaLinux 9.5 configuration and dual-boot recommendations
• Installation and optimization of Visual Studio Code (VS Code) for remote development

💻 Remote Access Configuration (ui-ams & CNAF) Obtained active CNAF and CERN accounts and verified access credentials.

Configured secure connections to the bastion.cnaf.infn.it server and the ui-ams virtual machine using SSH with ProxyJump.

Set up passwordless login with a generated ed25519 SSH key pair and fallback RSA key.

Configured .ssh/config file to streamline connections using VS Code and terminal sessions.

🖥 AlmaLinux 9.5 Development Environment Installed AlmaLinux 9.5 in a dual-boot configuration to enable high-performance native compilation of C++ code using NAIA and NSL.

Installed critical packages including:

• Development Tools
• ROOT-compatible libraries
• HTCondor clients
• Deployed Visual Studio Code locally and validated remote editing via the Remote - SSH extension.

🛠 Local-Remote Integration and Testing Used the guide's steps to:

• Edit and debug NAIA code on ui-ams via VS Code.
• Run initial test jobs on HTCondor with ROOT input files.
• Organize project directories across local AlmaLinux and CNAF environments.

📞 Zoom Meeting with Dr. V. Formato – Definition of Data Analysis Targets

A Zoom call was held with Dr. V. Formato (INFN TOV) to define the scientific focus of the data analysis within the NAIA framework. It was agreed that the primary objective would be the identification and characterization of key components of space radiation in the low-energy range (<10 GV), which are crucial for space radiobiology applications.

In particular, the initial phase of the analysis will focus on time-resolved variations in proton fluxes during the period January–February 2022, a timeframe known for notable solar energetic particle (SEP) activity. This will serve as a reference for identifying biologically significant radiation events relevant to short-duration space missions and astronaut exposure scenarios.

🛠️ Installation and Configuration of the NAIA Framework During April, technical work focused on setting up the software environment necessary to begin the AMS data analysis, particularly on the installation and configuration of the NAIA and NSL libraries. These tools are essential for the processing and interpretation of AMS proton data in the context of space radiobiology.

The NAIA framework was successfully set up following the official documentation (v1.2.0). The installation was carried out and tested both on a local development machine and on the CNAF infrastructure (ui-ams), where all necessary dependencies are pre-installed and maintained. Compatibility with CentOS7 and EL9-based systems was confirmed.

Key components of the setup included:

• Proper linking of the NAIA and NSL libraries for the analysis project.
• Organization of the project directory to include source files, build directories, and required environment setups.
• Verify that the analysis framework integrates with CMake to properly compile and execute the analysis chain.
• The NSL library, which is required for applying selection criteria and corrections in the NAIA chain, was also successfully integrated.

📊 Initial Data Analysis: Daily Proton Fluxes (Jan–Mar 2022)

Alongside the setup activities, the first steps of data analysis were initiated, focusing on daily variations in proton fluxes for the time window January to March 2022. This period was chosen due to known solar activity relevant to short-term radiobiological effects in space environments. The identification of key radiation components in the low-energy range (<10 GV) was defined as a primary scientific objective. These components—such as protons, alpha particles, and heavy ions (e.g., iron nuclei)—are essential for modeling acute exposure scenarios relevant to space radiobiology applications, including solar particle events (SPEs). The workflow for this analysis will follow the methodology developed by F. Faldi et al., which includes:

• Event selection procedures
• Efficiency corrections
• DailyProton flux computation and normalization

This structured approach ensures that the data are filtered, corrected, and interpreted consistently for biological modeling purposes. The analysis pipeline has been tested using small data samples and will be scaled to larger sets using HTCondor batch processing on CNAF in the following months.

This activity is directly connected to the 25th March 2025 Zoom training meeting about analyzing Proton Daily fluxes data from the AMS raw data, organized by V. Formato and held by F. Faldi. The codes developed by F. Faldi et al were used as a starting point.

📌 May 2025 – Week 1

🧪 Summary of Activities:

Regenerated output files from the standard and Monte Carlo analyses due to a previously undetected typo in selector.cpp. The corrected files (data.root and mc.root) were shared with Francesco and successfully verified.

Attempted to run assembler.py on AlmaLinux, but experienced instability across multiple Python versions and library configurations. Ultimately, the script was run on Windows 11 to ensure stability.

Processed daily proton flux data using assembler.py for the period from 1 January to 1 March 2022 (days 3880–3938). Output was successfully generated for all days except day 3931. Results (including .txt and .png files) were uploaded to Pandora Cloud as files.zip.

Contributed to documentation efforts by completing guideline sections on the INFN wiki.

📓 May 2025 – Week 2

🧪 Activities: Since the reference proton flux data currently extends only up to 2019 (an update to 2024 is expected soon from Francesco), the existing daily_compiler.py script was limited to plotting data only up to 2019. To proceed with our 2022 analysis, I modified the script to allow plotting of proton flux data even in the absence of reference values beyond 2019.

After modifying the script, I successfully plotted the daily proton flux data for the period 1 January to 1 March 2022. The figures were sent in a previous email.

Completed the "Technical Guide" on the INFN wiki. It includes updated procedures and relevant modifications. Review and feedback are welcome to ensure completeness and accuracy.

📓 May 2025 – Week 3

🧪 Activities: Shared the newly generated daily flux data with Francesco. Upon review, he noted unusually high statistical errors in the flux values. He re-examined his script (assembler.py) and identified a mistake in the error calculation block.

After correcting the script, he updated it in the CERN repository. I then regenerated all daily flux data files using the revised version of assembler.py. The updated files are available in the daily_Flux_Data.zip archive and were sent via email.

Requested Francesco’s daily flux data for comparison. Since the AMS reference data was limited to 2019, his updated 2022 dataset was critical for validating our results. He uploaded his ROOT data file to the CERN repository and also modified daily_compiler.py to allow side-by-side comparison of user and reference datasets.

Upon testing the modified script, found that the figures were plotted incorrectly due to mismatches between the code and the structure of the uploaded ROOT file. I corrected the script and shared the fix with Francesco, who subsequently updated both the script and the ROOT file in the repository.

📓 May 2025 – Week 4

🧪 Activities: Work on executing the assembler.py script using HTCondor to enable parallelized data processing. Unlike the C++ code (selector.cpp), Python scripts do not require compilation, making their execution on HTCondor similar to running bash scripts.

The HTCondor execution process for assembler.py still requires optimization. Once finalized, Full procedure documented on the INFN Wiki to complete the current technical guide.

📁 Files Generated: data.root, mc.root – regenerated and validated output files.

Daily proton flux text and plot files (.txt, .png) stored in files.zip on Pandora Cloud.

⚙️ Tools Used: C++ (selector.cpp)

Python scripts: assembler.py, daily_compiler.py

Operating Systems: AlmaLinux (unstable), Windows 11 (stable execution)

Pandora Cloud for output storage

🚧 Challenges Encountered: Instability of assembler.py on AlmaLinux, despite testing across Python/library versions.

Execution failure for day 3931 in the proton flux series (issue under investigation).

📌 Planned Next Steps: Run daily_compiler.py to visualize and compile daily proton flux data.

Continue refining documentation on the INFN wiki.