News

This article first appeared on The Daily Sun on 15 January 2023 and is available here. Another version in Bengali appeared in The Daily Ittefaq and is available here.

Bangladesh has the largest river delta in the world, which means much of the landscape is composed of alluvial floodplains. The northeast and southeast regions rise into small hills, and its river systems give life to coastal areas, wetlands, forests, fisheries, tea plantations, and industries. Within this great ecological diversity, climate change impacts and rapid land use and land cover changes have led to growing environmental concerns around water availability, siltation and erosion, groundwater contamination and salinity, increasing incidences of cyclones and flooding, and forest degradation.

Bangladesh is accordingly expanding its information technology capabilities to carefully govern and manage its natural resources. It is increasingly using remote-sensing technology and satellite imagery for natural resource management, development projects, and disaster management. For example, the Bangladesh Forest Department uses remote-sensing technology in land cover mapping for forest inventory, forest change detection for forest management, and national and international reporting. Similarly, the Department of Agricultural Extension is mapping crop patterns and damage using publicly available Sentinel-1 SAR and Sentinel-2 optical images. Other departments also use satellite images and remote-sensing technology for road, highway, and bridge construction and management; agricultural planning; and space research, among other fields.

The era of satellite remote sensing began when the former Soviet Union launched Sputnik 1, Earth’s first artificial satellite, on 4 October 1957. From then, there has been no end to the massive human interest in artificial satellites and their images. Earth-observing satellites have provided iconic views and unprecedented scientific insights.

However, many public and private organisations in Bangladesh appear to prefer using very high-resolution commercial satellite images to conduct projects. Many professionals – i.e. decision makers and analysts – seem to assume that acquiring expensive, high-resolution images will make mapping better. There are, of course, many apparent advantages of high-resolution satellite images, but purchasing these expensive commercial images may not always be the best approach in satellite-based mapping approaches. In practice, the accuracy of satellite image mapping depends on satellite image analysis skills, as well as the image analysis software and algorithms used. If the remote-sensing analyst is skilled in image processing and extracting information from satellite data, it is possible to use freely collected satellite images to create maps of almost equivalent quality to purchased satellite images, as done in this study on natural coastal expansion in Bangladesh which uses publicly available annual composite data.

Very high-resolution satellite images are exorbitantly priced (at around USD 15 per square kilometre on average), making it impossible to purchase as many images as desired in any given month or year. Since most high-resolution satellites have been launched only in recent years and multispectral image archives (i.e. colour images suitable for mapping) did not exist before 2000, high-resolution satellite imagery cannot help us analyse long-term changes from the past. Another disadvantage in purchasing any commercial satellite image is the prolonged procurement process. Free satellite images, in contrast, have a wider temporal reach and can be acquired anytime. And if a project requires very high-resolution images urgently, drone multispectral images (which are cheaper than commercial satellite images) can be used with the government’s permission.

Therefore, government departments and organisations in Bangladesh that work on satellite-based maps, including the Bangladesh Space Research and Remote Sensing Organization, could consider the advantages of prioritising the wider use of free satellite images. In doing so, it could save its foreign currency reserves and divert resources towards building the capacity and skills of human resources. For example, the country can focus on building competency in satellite remote-sensing data analysis using Google Earth Engine, an online endeavour by Google that can help process several petabytes of public imagery datasets, including Landsat, Sentinel, and MODIS. Google Earth Engine runs on Google Cloud, giving access to high-performance computing (even from a mobile device) and functioning well even on moderate internet bandwidth. This is especially useful for analysts working in countries with poor or unstable internet connections.

Moreover, there is no dearth of options when it comes to publicly available satellite imagery. Earth observation Landsat satellites launched by the U.S. Geological Survey and NASA have been providing uninterrupted data since 1972 to make informed decisions about natural resources and the environment. Since then, many satellites have been launched by governments and private organisations for commercial and non-commercial purposes to acquire Earth images.

The free distribution of Landsat satellite images to all users worldwide in late 2008 marked another breakthrough for scientific use. Copernicus Sentinel images were made free and open to the public right from launching of the Copernicus Sentinel Missions series satellites in 2014. Copernicus Sentinel images are useful for monitoring river and river and coastal courses; mapping inundation; measuring earthquake damage; and determining changes in glaciers, snow, and land cover and land use.

Government agencies in Bangladesh can therefore use such publicly available satellite images for various planning and management initiatives, including natural resource management, natural disaster mapping and monitoring, mapping of different croplands patterns, and monitoring and measuring the status of forests and other vegetation.

Author