Changes in agricultural crop suits and landscape: a time-series study using multispectral imageries from rural Bangladesh

Abstract

Bangladesh is predominantly an agricultural country, so studying suitable land for crops and landscapes is essential. Substantial or detailed studies, particularly on this aspect of the landscape and crop suitability for different seasons, are limited. In this paper, an attempt has been made to analyze and evaluate the chronological changes in agricultural crop suits and landscapes along the river Titas, Bangladesh. For the study, the Brahmanbaria district was considered as a case study period from 2012 to 2022. Primary data were obtained from 384 Questionnaire surveys and 10 FGDs, whereas secondary data were collected from CEGIS and BBS. Sentinel 2 images were used for mapping and chronological changes in Land Use and Land Cover (LULC). Between the study periods, paddy cultivation declined from 31 to 18% in Rabi, leading to maximum winter vegetation. However, there were no significant changes in the Kharif-I season. Except for paddy farming, which declined from 22% to 18% in Kharif-II, changes are few. This study identified 12 driving variables demonstrating that labor and capital availability were more responsible, whereas farm size was less responsible for determining agricultural and landscape changes. Rabi, Kharif-I, and Kharif-II had an overall accuracy of 85.98, 82.13, and 89.52%, respectively, with a kappa coefficient of 84.90, 86.37, and 82.50%. All classes were over 83.87% in producer accuracy, while all classes except urban and industrial were above 86.00% in user accuracy.

J. Asiat. Soc. Bangladesh, Sci. 49(1): 85-100, June 2023