Mapping the Taliban takeover of Afghanistan

2022

Mapping the Taliban takeover of Afghanistan

Country/area: United Kingdom

Organisation: BBC News

Organisation size: Big

Publication date: 9/7/2021

Credit: Sayed Abdullah Nizami, Meena Baktash, Ismael Sadaat, Dominic Bailey, Mike Hills, David Brown, Johannes Dell, Roberto Belo Rovella, Ana Lucia Gonzalez, Sean Willmott, Gerry Fletcher, Mark Bryson, Bella Hurrell.

Biography: The BBC Afghan Service broadcasts in Pashto and Dari for audiences in Afghanistan and Pakistan on television, radio and online. The BBC News Visual journalism team is made up of journalists, designers and developers who engage digital BBC audiences through visually-led storytelling. Using compelling content and interactivity we focus on telling the most important stories and making complicated information easier to understand. The wider team encompasses journalists across the BBC World Service.

Project description:

The BBC Afghan Service gathered original data on the Taliban advance across Afghanistan and the Visual Journalism team mapped it on an almost daily basis from 9 July – 16 August, showing how districts and cities fell to the insurgents. The maps were widely used on the BBC’s UK and world news channels, news online, social media output, and about 40 World Service language sites and services.

Impact reached:

The maps we made with original data enabled us to tell the unfolding story of the Taliban advance in a visual and timely way, across a range of platforms, providing maximum value for our audience. We used the maps to tell the story of the how districts and provinces initially flipped back and forth between government and Taliban forces in July and then, as momentum built, the relentless advance of the Taliban in the second week of August. The last map we did was on 16 August – the day after Kabul fell and President Ghani had fled – by this time the only area left free of Taliban control was a small part of the Panjshir valley. The maps were used widely in news stories, features and live pages on the website over a six week period, as well as by the BBC’s World and UK News channel and the main TV bulletins on BBC One – the Six and the Ten. We published many of the maps together in our regularly updated Mapping the Taliban feature, which got several million page views. They also featured on social media and nearly all 40 World Service Languages used at least one version of the maps in their coverage.

Techniques/technologies used:

In Afghanistan group of journalists from the BBC Afghan Service took on the job of keeping across all the 398 districts and verifying reports of fighting and battles for control. The data was collated and input by a senior journalist in Kabul. The Afghan service set out the criteria for each control category . A district was classified as either Taliban or Government depending on who controlled the administrative centre, police headquarters and other government institutions. Regional reporters sent information, and that was supplemented with additional info from Taliban and government contacts, as well as a variety of local sources. In the UK we set up the spreadsheet for our Kabul colleagues, which they could update using agreed protocols, so we would know when changes had been made to each district. A CSV downloaded from the spreadsheet and was used to create a map in QGIS using a 2005 Afghanistan government boundaries shapefile that divided Afghanistan into 398 districts. The svg we output from QGIS was then passed to a designer to style and output the final png file in English and multiple other languages for publication online or use on TV output. This happened each time an update was needed. We used the data to create single maps, side by side comparisons, zoomed in maps to show change detail and an animated sequence to show change over time. We shared files with TV and language service colleagues.

What was the hardest part of this project?

When we considered how to tell the story of Afghanistan in early July – as the US and Nato forces left and the Taliban immediately started to take over parts of the country – we wanted a reliable and timely source for mapping control because it was a crucial part of the story. The challenge we faced was that other sources which were widely used by other news orgs, were updated less frequently than we would wish. The head of the BBC Afghan service was also keen for her team to give an accurate picture of the changing state of control using their expertise and journalism. By far the hardest aspect was the fact that our colleagues working in Afghanistan were gathering and verifying data under incredibly difficult circumstances. They continued to work despite the huge political and social upheavals taking place in their country. By early August only one extremely dedicated Afghan colleague remained on the project – and he sent us updates at all hours, sometimes several times a day, right up until the day after Kabul – the city he was living in – was invaded by the Taliban. This is an effective and successful mapping project but it also deserves recognition because of the professionalism and commitment to original journalism from our colleagues in Afghanistan.

What can others learn from this project?

Great projects can emerge from collaborations! We are so lucky at the BBC to have colleagues who work around the globe and bring unique skills and experience. But it really can be effective to reach out and work with different teams both inside and outside your organisation to accomplish projects you could never do as a single team. Simple tools can create very useful and effective assets but do consider how sustainable your workflow might be. The maps we made were produced manually each time and we didn’t script the production of them, because at the outset we did not anticipate the story moving so quickly. In the event, the way we produced the maps meant that they were easier for the BBC’s World Service websites to translate into up to 40 languages, but if the project has been sustained over a longer time then we might have found a more automated workflow.