Forecasting Sugarcane Area, Production and Productivity in Uttarakhand, India, up to 2029-30: A Rolling-origin Comparison of Time-series Models

Seema Arya *

Department of Agricultural Economics, ICAR-IARI, New Delhi, Delhi, 110012, India.

Ajay Kumar Tripathi

Department of Agricultural Economics, GBPUAT Pantnagar, Rudrapur, Uttarakhand, 263145, India.

*Author to whom correspondence should be addressed.


Abstract

This study examined trends in sugarcane area, production and productivity in Uttarakhand and its Kumaon and Garhwal divisions and developed forecasts to 2029–30. The analysis used 26 crop seasons (2000–01 to 2025–26) from the Directorate of Economics and Statistics, audited against the state Sugarcane Department series. Thirteen candidate procedures, comprising benchmark, trend-regression, exponential-smoothing, ARIMA and combination models, were compared using rolling-origin out-of-sample evaluation. A structural break around 2018–19 was assessed and treated through a level-shift adjustment. Production forecasts were derived from area and productivity, while divisional area forecasts were reconciled with the state forecast. State sugarcane area declined by 1.27% annually (P < .001), whereas productivity increased by 1.61% annually (P < .001), with much of the productivity increase occurring as a step change around 2018–19. State production showed no significant trend (P = .35). For 2029–30, the state forecasts were 79.8 thousand ha for area (95% prediction interval: 69.3–90.2), 6,025 thousand t for production (4,703–7,347) and 75.5 t ha⁻¹ for productivity (60.0–91.0). In the post-break nested holdout, the selected procedures did not outperform the naive forecast for eight of the nine series. Divisional forecasts were also sensitive to the structural break and inconsistencies in the source data. Accordingly, the forecasts are best interpreted as scenario ranges, and the principal contribution is the comparative evaluation framework.

Keywords: Sugarcane, forecasting, ARIMA, exponential smoothing, rolling-origin validation, structural break, forecast reconciliation, conformal prediction, Uttarakhand, Kumaon, Garhwal


How to Cite

Arya, Seema, and Ajay Kumar Tripathi. 2026. “Forecasting Sugarcane Area, Production and Productivity in Uttarakhand, India, up to 2029-30: A Rolling-Origin Comparison of Time-Series Models”. Journal of Scientific Research and Reports 32 (11):1440-52. https://doi.org/10.9734/jsrr/2026/v32i114764.

Downloads

Download data is not yet available.