Stolen Strata: Quantifying the Anthropogenic Erasure of Kashmir's Karewa Terraces and Its Threat to the Saffron Economy
Abstract
Karewa tablelands are the flat-topped, scarp-bounded remnants of the old lake and river deposits of the Kashmir Valley. Reporting from Kashmir has said for years that they are being dug away for brick clay and fill, but I did not find a satellite measurement of how much, where and when. This study asks three questions: how much karewa tableland that was vegetated in 2013–2015 is no longer vegetated in 2023–2025; what that land has become, in particular how much of it now lies inside brick-kiln fields; and where and when the change happened. I map 180 scarp-bounded tablelands (173.3 km²) in a 2,160 km² box across Budgam, Pulwama and Srinagar districts from the Copernicus elevation model, using slope, height above drainage and the share of each flat top's edge that is a scarp. I then follow the yearly peak NDVI of every 30 m pixel in Landsat 8/9 alone and flag land that was clearly vegetated in 2013–2015 and is not in 2023–2025, with a strict and a looser test. They flag 107 ha and 329 ha of the tablelands. Two stratified random samples, 240 points in all, labelled blind to stratum against recent imagery, show that about 299 ha (90%) of the flagged land is not vegetated today and that about 262 ha of it was vegetated in imagery of 2013–2015. Under a written rule for where a kiln field ends, applied afterwards to the edge points of both samples (this relabelling was not blind to stratum), about 210 ha (95% interval roughly 189–231 ha) lies inside brick-kiln fields, 197 ha of it on land that the older imagery also shows as vegetated. The change is concentrated in two kiln belts in Budgam. At Rangeen Kultreh a kiln field opens after 2016 on land that was vegetated in every usable year since 1993 but one anomalous year, and a pollution-control filing before the National Green Tribunal independently dates a kiln there to 2017. On the Bandagam–Batapora tablelands land with little vegetation was already present in the 1990s, and the looser test finds more post-2013 loss there than at Rangeen Kultreh. Outside these two belts the tablelands convert at the same rate as comparable flat, raised land. I could not measure how deep the ground was dug, and I found no conversion of this kind on the saffron tablelands of Pampore. An earlier version of this work reported that the bare-earth share of karewa terraces rose from 1.84% in 1994 to 8.43% in 2025. Section 4.1 shows that the rise came from comparing Landsat with a differently built Sentinel-2 product (on Landsat alone the 2025 figure is 0.07%), and it is withdrawn.
Related
- Reconstructing a 140-Year Historiography of Misinterpretation: A Forensic Reassessment of the Tarawera Eruption and the Pink and White Terraces
- Unreported mass movement cascades and future Glacial Lake Outburst Flood (GLOF) hazard in Jammu and Kashmir, Western Himalaya
- Anthropogenic groundwater declines may exert far-reaching climatic impacts
- Spatial modelling of cetacean habitats and cumulative anthropogenic pressures in the Caribbean: implications for transboundary marine governance
- Synergistic effects of climatic drivers and anthropogenic land degradation on cascading geomorphic hazards in Iran
BibTeX
@misc{maske2026stolen,
title = {{Stolen Strata: Quantifying the Anthropogenic Erasure of Kashmir's Karewa Terraces and Its Threat to the Saffron Economy}},
author = {SAKSHI D. MASKE},
howpublished = {EarthArXiv},
year = {2026},
month = sep,
doi = {10.31223/x5hb8j},
url = {https://eartharxiv.org/repository/view/14805/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Stolen Strata: Quantifying the Anthropogenic Erasure of Kashmir's Karewa Terraces and Its Threat to the Saffron Economy
AU - SAKSHI D. MASKE
PY - 2026
DA - 2026/09/04
PB - EarthArXiv
DO - 10.31223/x5hb8j
UR - https://eartharxiv.org/repository/view/14805/
AB - Karewa tablelands are the flat-topped, scarp-bounded remnants of the old lake and river deposits of the Kashmir Valley. Reporting from Kashmir has said for years that they are being dug away for brick clay and fill, but I did not find a satellite measurement of how much, where and when. This study asks three questions: how much karewa tableland that was vegetated in 2013–2015 is no longer vegetated in 2023–2025; what that land has become, in particular how much of it now lies inside brick-kiln fields; and where and when the change happened. I map 180 scarp-bounded tablelands (173.3 km²) in a 2,160 km² box across Budgam, Pulwama and Srinagar districts from the Copernicus elevation model, using slope, height above drainage and the share of each flat top's edge that is a scarp. I then follow the yearly peak NDVI of every 30 m pixel in Landsat 8/9 alone and flag land that was clearly vegetated in 2013–2015 and is not in 2023–2025, with a strict and a looser test. They flag 107 ha and 329 ha of the tablelands. Two stratified random samples, 240 points in all, labelled blind to stratum against recent imagery, show that about 299 ha (90%) of the flagged land is not vegetated today and that about 262 ha of it was vegetated in imagery of 2013–2015. Under a written rule for where a kiln field ends, applied afterwards to the edge points of both samples (this relabelling was not blind to stratum), about 210 ha (95% interval roughly 189–231 ha) lies inside brick-kiln fields, 197 ha of it on land that the older imagery also shows as vegetated. The change is concentrated in two kiln belts in Budgam. At Rangeen Kultreh a kiln field opens after 2016 on land that was vegetated in every usable year since 1993 but one anomalous year, and a pollution-control filing before the National Green Tribunal independently dates a kiln there to 2017. On the Bandagam–Batapora tablelands land with little vegetation was already present in the 1990s, and the looser test finds more post-2013 loss there than at Rangeen Kultreh. Outside these two belts the tablelands convert at the same rate as comparable flat, raised land. I could not measure how deep the ground was dug, and I found no conversion of this kind on the saffron tablelands of Pampore. An earlier version of this work reported that the bare-earth share of karewa terraces rose from 1.84% in 1994 to 8.43% in 2025. Section 4.1 shows that the rise came from comparing Landsat with a differently built Sentinel-2 product (on Landsat alone the 2025 figure is 0.07%), and it is withdrawn.
ER -