Analysis of historical dataset based on Rennell’s map on Ganges River basin

Abstract

Rennell has been widely known as the father of the modern maps since the publication of his survey works of 1760s. He made the first nearly accurate map of the then Bengal of the Indian Subcontinent entitled “A Bengal Atlas (1779)”, a work which was considered as the most significant geographical contributions of the British Empire for their strategic and administrative interests. The most important feature of the Atlas is the Ganges Basin which flows as the trans-boundary River of South Asia entrenching India, China and Bangladesh. This study reveals the historical and contextual phenomena of the river like old channels, onomastic and old administrative footprints from Rennell’s Atlas around the old courses of Ganges River, more specifically Plate 18 confining within the territorial extent of current Bangladesh Boundary entrapping the history more than 250 years of the aerial extent encompassing 11,651 km². This research also reveals the map accuracy by calculating the modern graticules comparing the former geographic coordinate systems. In order to develop historical dataset based on Rennell’s map, various modern tools of renowned GIS approaches and corresponding attributes like geo-referencing, digitizing, geo-corrections, projection transformation, calculating geometry, pivot table etc. In the 1760s, Rennell was dependent on the primitive sextant instrument and calculated latitude, longitude assuming Fort William of Kolkata as Zero (0) degree. In this research we integrated Bangladesh Transverse Mercator (BTM) considering World Geodetic System 1984 as datum. The results illustrate the volume of rivers, riverine forest, jheel, char land are occupying 116,600; 215,959; 3573; and 15,471 ha respectively in the Rennell’s map which are currently almost disappeared. In the said plate, marsh land covered the maximum area about 118,676 ha while length of Kaccha (unmetalled) road was 1339 km, though recently metalled road length has been increased dramatically. The detected broad administrative unit was 337. The error result in the present research is with 99% accuracy. This research has highlighted the strengths to retrieve underlying evidences and relevant features and statistical/attribute data from the historical maps using remote Sensing (RS) and Geographic Information System (GIS) approaches.