Ioniță et al. (2026) Spatiotemporal Variability of Seasonal Snow Cover over 25 Years in the Romanian Carpathians: Insights from a MODIS CGF-Based Approach

Identification

• Journal: Remote Sensing
• Year: 2026
• Date: 2026-02-02
• Authors: Andrei Ioniță, Iosif Lopătiță, Florina Ardelean, Flavius Sîrbu, Petru Urdea, Alexandru Onaca
• DOI: 10.3390/rs18030468

Research Groups

• Institute for Advanced Environmental Research (ICAM), West University of Timișoara, Romania
• Department of Geography, West University of Timișoara, Romania

Short Summary

This study analyzed 25 years of MODIS satellite data to map and quantify spatiotemporal changes in seasonal snow cover phenology across the Romanian Carpathians, revealing a significant shortening of the snow season primarily due to earlier melt at lower and mid-elevations.

Objective

• To provide the first comprehensive, multi-decadal assessment of snow cover phenology (duration, onset, end date) across the entire Romanian Carpathians (2000–2025) using MODIS satellite data, quantifying its spatial distribution, interannual variability, and long-term trends in relation to topography and climate.

Study Configuration

• Spatial Scale: Romanian Carpathians, covering approximately 66,303 square kilometers, with a spatial resolution of 500 meters.
• Temporal Scale: 25 years (2000–2025), using daily observations aggregated by hydrological year (1 October–30 September).

Methodology and Data

• Models used:
  • Normalized Difference Snow Index (NDSI) for snow detection.
  • Cloud-Gap-Filled (CGF) procedure for data continuity.
  • Mann–Kendall test for monotonic trend detection.
  • Theil–Sen slope estimator for trend magnitudes.
  • Ordinary least squares regression for Snow-Covered Area (SCA) trends.
  • F1 score for daily snow occurrence agreement.
  • Pearson correlation for Snow Cover Duration (SCD) agreement.
  • Spearman rank correlation for climate associations.
• Data sources:
  • MODIS/Terra MOD10A1F product (Cloud-Gap-Filled Snow Cover Daily L3 Global 500 m SIN Grid, Version 6.1) from NASA EarthData via NSIDC-DAAC.
  • FABDEM (Forests and Buildings Removed Digital Terrain Model) at 30 m spatial resolution.
  • MeteoManz (SYNOP and BUFR observations) from 26 meteorological stations for snow depth validation.
  • Homogenized Romanian climate dataset for seasonal air temperature and precipitation (43 meteorological stations, 2000–2023).
  • Gridded E-OBS temperature and precipitation fields (9 km).

Main Results

• The average Snow Cover Duration (SCD) across the Romanian Carpathians for 2000–2025 was 48.2 days (median 39.9 days).
• The mean Snow Onset Date (SOD) was day 82.8 ± 15.8 of the hydrological year, and the mean Snow End Date (SED) was day 151.9 ± 8.3.
• The snow season shortened by 1.14 days per year (approximately 11 days per decade) across the entire range, primarily driven by earlier melt (−0.78 days per year, statistically significant) rather than a uniform delay in onset (+0.28 days per year, non-significant).
• Strongest SCD reductions (ranging from −17 to −25 days per decade) occurred at low and mid-elevations (<1000 meters).
• High-elevation areas (>2000 meters) exhibited considerably more stable snow regimes, with minor changes in SCD (−0.13 days per year), SOD (+0.46 days per year), and SED (+0.32 days per year), and SCD commonly exceeding 180–200 days.
• North-facing slopes consistently showed longer SCD (10–12 days more than south-facing slopes at most elevation bands).
• Snow-Covered Area (SCA) exhibited declining trends across all three Carpathian units (Western, Eastern, Southern) and seasonal windows (e.g., winter SCA decreased by −0.936% per year in the Western Carpathians).
• Snowline Elevation (SLE) displayed a pronounced seasonal cycle with the greatest interannual variability during core winter and early spring (December–March).
• Validation against station data showed strong agreement for MODIS-derived SCD (Pearson r = 0.95, Mean Absolute Error = 16.7 days, median bias = −8.0 days).
• Seasonal air temperature was identified as the dominant climatic covariate of regional snow phenology variability, with cold-season temperature increasing by +0.64 °C per decade, while precipitation associations were weaker.

Contributions

• This study provides the first comprehensive, multi-decadal (25-year), spatially continuous assessment of snow cover phenology across the entire Romanian Carpathians using MODIS Cloud-Gap-Filled data.
• It closes a significant data gap for a climatically distinct and comparatively under-instrumented mountain region in Southeastern Europe.
• The research establishes a robust baseline for future cryospheric, ecological, hydrological, and periglacial research, including investigations of snow–ground thermal coupling and the sensitivity of periglacial processes to snow variations.

Funding

• Ministry of Research, Innovation and Digitization through CNCS–UEFISCDI, project PN-IV-P2-2.1-TE-2023-0603 (PNCDI IV framework).
• UEFISCDI, project PN-IV-P6-6.3-SOL-2024-0248, contract no. 17SOL (T17).

Citation

@article{Ioniță2026Spatiotemporal,
  author = {Ioniță, Andrei and Lopătiță, Iosif and Ardelean, Florina and Sîrbu, Flavius and Urdea, Petru and Onaca, Alexandru},
  title = {Spatiotemporal Variability of Seasonal Snow Cover over 25 Years in the Romanian Carpathians: Insights from a MODIS CGF-Based Approach},
  journal = {Remote Sensing},
  year = {2026},
  doi = {10.3390/rs18030468},
  url = {https://doi.org/10.3390/rs18030468}
}