Remote Sensing Laboratory University at Buffalo Mohammad Salmani
(a) Then: illustration of a tidewater glacier whose ice cliff meets a fjord of floating ice between snow-streaked mountains.

PhD Student Earth Sciences

Mohammad Salmani

I map how the Greenland Ice Sheet has retreated since the Little Ice Age, using satellite imagery, historical aerial photographs and machine learning.

Skip to About

Fig. 1Frame 1 of 5

Tidewater glacier retreat — an illustrative sequence, not observational data.

Scroll to advance the sequence

About

I am a PhD student in the Department of Earth Sciences at the University at Buffalo, working in the Remote Sensing Laboratory (opens in new tab) with my advisor, Beata Csatho.

My PhD research reconstructs the Greenland Ice Sheet since the Little Ice Age: where its margin stood at its maximum, how it has retreated from there, and how the surface elevation of its glaciers has changed.

Research interests

  • Remote sensing of the cryosphere and ice-sheet change
  • Reconstruction of Greenland Ice Sheet retreat since the Little Ice Age
  • Machine learning and deep learning for glacier and ice-sheet mapping
  • Surface-elevation change from historical aerial photogrammetry and satellite altimetry
  • Implications of ice-sheet change for sea-level rise and climate
Portrait of Mohammad Salmani, smiling, in a black T-shirt.

Education

  • –present PhD, Earth Sciences, University at Buffalo (SUNY).
  • B.S., Surveying and Geospatial Engineering, University of Tehran.

Research experience

  • –present Graduate Research Assistant, Remote Sensing Laboratory.
  • Research Intern, Geodesy Lab, University of Tehran. GNSS/INS integration in MATLAB.

Teaching & mentoring

  • REU Mentor. Channel mapping at Eqip Sermia, Greenland, from SAR imagery in Google Earth Engine and QGIS.
  • Teaching Assistant, Principles of Database, University of Tehran, spring and fall.

Awards & honors

  • Pegrum Professional Development Award, Department of Earth Sciences, UB
  • Graduate Student Association Conference Funding, UB,
  • Ranked 1st in cumulative GPA in the major, University of Tehran
  • National Undergraduate Full Scholarship, University of Tehran
  • Exemption from the M.Sc. National Entrance Exam (Iran) as an exceptional talent

Research

Two lines of work, from archival aerial photographs to machine learning on satellite imagery.

Greenland Ice Sheet retreat since the Little Ice Age

I map Greenland’s Little Ice Age ice margins and trimlines in Sentinel-2 and Landsat imagery, processed in Google Earth Engine, to track how the margin has retreated through time. I also reconstruct surface-elevation change of Greenland glaciers from historical aerial photographs.

Data
Historical aerial photographs, Sentinel-2, Landsat

Finding

Of the ice area Greenland has lost since the Little Ice Age, nearly half has gone since 2000.

Salmani et al. · preprint, in review

Machine learning for mapping ice extent

I develop machine-learning approaches that map glacial trimlines and Little Ice Age maximum glacial extents automatically from Sentinel-2 imagery. Earlier, I co-authored a study of how transferring pre-trained weights affects a Siamese deep-learning network for change detection.

Data
Sentinel-2
Methods
Machine learning, deep-learning change detection

Tools

Python, MATLAB, R; Google Earth Engine, ArcGIS, QGIS, ENVI, Agisoft Metashape

Publications

In review

  1. 2026

    M. Salmani, B. Csatho, J. Briner, S. Nowicki, I. Parmuzin, H. Eberhardt, H. Gao, A. C. Luzardi, G. Talebigheshlaghi. Nearly half of Greenland’s post-Little Ice Age area loss occurred since 2000. In review, Communications Earth & Environment (2026).

    doi:10.21203/rs.3.rs-9125064/v1 (opens in new tab) Preprint, Research Square (opens in new tab)

In preparation

  1. M. Salmani, B. M. Csatho, J. P. Briner, S. Nowicki, I. Parmuzin, A. C. Moraes Luzardi. Machine Learning Approaches for Automated Mapping of Glacial Trimlines Using Sentinel-2 Imagery.

  2. A. C. M. Luzardi, S. Nowicki, B. Csatho, J. Briner, M. Salmani, H. Gao, D. Felikson, J. Cuzzone, K. Poinar. Reconstructing Greenland Ice Sheet Elevation in 1850.

Published

  1. 2023

    M. Aghdami-Nia, R. Shah-Hosseini, M. Salmani. Effect of Transferring Pre-trained Weights on a Siamese Change Detection Network. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, X-4/W1-2022, 19–24 (2023).

    doi:10.5194/isprs-annals-X-4-W1-2022-19-2023 (opens in new tab) Article (opens in new tab)

Full list on Google Scholar (opens in new tab)

Presentations

2025–26

  1. Poster · AGU Fall Meeting 2025

    Temporal Evolution of Greenland Ice Sheet Retreat Since the Little Ice Age

    C21F: Remote Sensing of the Cryosphere: Glaciers and Ice Sheets III

    New Orleans, LAAbstract C21F-0908 (opens in new tab)

    Also presented

    (a) Mohammad Salmani pointing at his poster, which shows four colored maps of Greenland, while a visitor looks on. (opens a larger version) (b) Mohammad Salmani, holding a cup, in front of his poster “Temporal Evolution of Greenland Ice Sheet Retreat Since the Little Ice Age” in a gym set up for the poster session. (opens a larger version)
    Fig. 2Poster C21F-0908. (a) AGU Fall Meeting 2025, New Orleans. (b) Celebration of Student Academic Excellence, UB, April 2026.

  2. Poster · AGU Fall Meeting 2025

    A Machine-Learning Framework for Mapping Little Ice Age Maximum Glacial Extents

    C41G: Machine Learning in the Cryosphere III

    New Orleans, LAAbstract C41G-1100 (opens in new tab)

    Mohammad Salmani pointing at the satellite-image panels of his machine-learning poster at board 1100 while a visitor listens. (opens a larger version)
    Fig. 3At board 1100, AGU Fall Meeting 2025, New Orleans.

2024

  1. Poster · AGU Fall Meeting 2024

    Automated Mapping of Greenland Ice Sheet Trimlines Using Sentinel-2 Imagery

    C43D: Understanding the Drivers of Mountain Glacier and Ice Sheet Change Across Contemporary and Paleoclimate Timescales I

    Washington, DCAbstract C43D-0288 (opens in new tab)

    Mohammad Salmani, arms crossed, beside his poster “Automated Mapping of Greenland Ice Sheet Trimlines Using Sentinel-2 Imagery” at board 288. (opens a larger version)
    Fig. 4At board 288, AGU Fall Meeting 2024, Washington, DC.

Outreach & Events

Science communication

  • The Pulse of Greenland

    An artistic visualization of more than a century of Greenland ice-margin retreat, shown at the Art of Research Competition, Buffalo Museum of Science, Buffalo, NY.

    Mohammad Salmani smiling beside “The Pulse of Greenland”, a dark print of concentric radial bar rings for the periods 1850 to the 1980s, the 1980s to 2000, and 2000 to 2022. (opens a larger version)
    Fig. 5“The Pulse of Greenland” at the Buffalo Museum of Science. Photo: Cheryl Gorski

Citizen science

Conferences & workshops

Photo