default author photo

Mashrekur Rahman

Dartmouth College

SCHOLARLY PAPERS

5

DOWNLOADS

130

TOTAL CITATIONS

1

Scholarly Papers (5)

1.

Drought Awareness Over Continental United States

Number of pages: 47 Posted: 30 Mar 2024
Dartmouth College, University of California, Davis, University of California, Davis - Department of Land, Air and Water Resources, University of California, Davis, Government of the State of Israel and affiliation not provided to SSRN
Downloads 58 (989,921)
Citation 1

Abstract:

Loading...

droughts, Computer Vision, Google Trends, Twitter, SPEI, Machine learning

2.

Mini-JEPA Foundation Model Fleet Enables Agentic Hydrologic Intelligence

Number of pages: 47 Posted: 23 May 2026
Mashrekur Rahman
Dartmouth College
Downloads 27 (1,448,301)

Abstract:

Loading...

JEPA, Geospatial foundation models, Sensor specialization, Agentic retrieval, Hydrologic intelligence

3.

New Metrics for Characterizing and Comparing Soil Moisture Dynamics Across Depths and Locations

Number of pages: 33 Posted: 24 Feb 2026
affiliation not provided to SSRN, affiliation not provided to SSRN, University of Georgia, Dartmouth College and University of Georgia
Downloads 19 (1,491,251)

Abstract:

Loading...

Soil moisture dynamics, soil moisture recession slope, slope reversal frequency, soil moisture flashiness, and integrated soil moisture control

4.

Physically Interpretable AlphaEarth Foundation Model Embeddings Enable LLM-Based Land Surface Intelligence

Number of pages: 47 Posted: 13 Feb 2026
Mashrekur Rahman
Dartmouth College
Downloads 15 (1,548,648)

Abstract:

Loading...

Foundation models, AlphaEarth, Embedding interpretability, Remote Sensing, Large Language Models

5.

Drivers of Soil Moisture Dynamics over the Continental United States

Number of pages: 49 Posted: 05 May 2026
Dartmouth College, affiliation not provided to SSRN, affiliation not provided to SSRN and affiliation not provided to SSRN
Downloads 11 (1,548,648)

Abstract:

Loading...

soil moisture, Vision Transformers, Feature importance, SHAP, Continental United States, Deep learning