Nangarhar is the second-most-deadly province for civilians in all of Afghanistan, where the world's deadliest conflict rages on in its 18th year.
And the remote majestic mountains in that eastern province are the stronghold of extremist Islamic State (IS) fighters -- making the region a magnet for deadly militant bombings and U.S. air strikes.
Oddly enough, Nangarhar achieved the implausible feat of attracting the second-highest number of total votes in the war-torn country's September 28 presidential election, according to Afghanistan's Independent Election Commission (IEC).
The seemingly inflated number of votes and even turnout in Nangarhar -- along with improbable figures in other parts of Afghanistan -- has raised suspicions that the outcome of another landmark election could be tainted by fraud.
Election officials have hailed the recent election as the "most transparent" ever, following chaotic parliamentary elections in October 2018 and a fraud-marred presidential poll in 2014.
But allegations of fraud have surged as the IEC prepares to announce the long-awaited preliminary results on October 19.
The two front-runners in the election -- incumbent Ashraf Ghani and current Chief Executive Abdullah Abdullah -- have both claimed victory and signaled they will not accept defeat, a scenario that could trigger a political crisis and spill over into violence.
Observers say new rules introduced to combat fraud -- computerized voter lists and biometric voter verification -- prevented large-scale ballot stuffing but opened the process to new forms of manipulation.
New Rules, New Problems
The sophisticated biometric system included fingerprints and facial recognition. Additionally, voters could cast their ballots only in polling stations where they had registered to vote.
But the system caused its own problems. There were cases of biometric devices malfunctioning, vanishing, and untrained polling staff unable to use the machines.
Some devices were unable to find voters' names on voter lists, raising concerns of voter suppression. "Why were names missing from lists?" asks Srinjoy Bose, an international election observer who visited polling stations in the capital, Kabul, on election day. "Was it a system failure or were names scrubbed from the lists deliberately?"
Addressing the issue on election day, the IEC eased restrictions, allowing anyone with election stickers on their national identity cards to vote, paving another way for potential fraud.
That decision has created confusion and disputes over what votes should be counted or discarded, leading to the IEC declaring that only biometrically verified votes will be counted.
The campaigns of Ghani and Abdullah have sparred bitterly over the issue.
"The biometric system has merely deluded people into thinking that it has solved the problem of irregularities, but what it has done is actually created different problems," says Bose, an assistant professor at the University of New South Wales in Australia. "Candidates are likely to seize upon this issue, seriously calling into question the legitimacy of the process; even if large-scale fraud -- ballot stuffing, for example -- was not committed."
In a sign of a potential dispute, the Independent Electoral Complaints Commission (IECC) has received over 4,500 complaints. The commission will process and rule on them before preliminary results are released on October 19.
Thomas Ruttig, co-director of the independent Kabul-based Afghanistan Analysts Network, says that "everything now depends" on whether the IEC sticks to its original criteria -- counting only biometric-verified votes -- or "waters down" its conditions.
"The introduction of the biometric system, of course, makes fraud more difficult," Ruttig adds. "But people might try anyway, hoping they get away with it."
Videos purporting to show election workers committing fraud have been widely shared on social media. The authenticity of the videos could not be independently verified.
"It is highly possible that there was fraud," says Hameed Hakimi, a research associate at the London-based Chatham House think tank. "However, the nature of fraud associated with the biometric system will be different to the 'paper stuffing' that was seen in the previous presidential election."
Hakimi says Afghanistan's weak telecommunications infrastructure make the mobile networks vulnerable to Taliban attacks, and the biometric machines relied heavily on those networks.
Prior to the election, the Taliban targeted telecommunication towers in several parts of Afghanistan, blowing them up and disrupting mobile-phone services in districts and even some cities.
The IEC said it had no contact with around 900 of the approximately 4,700 polling stations across the country because of the attacks on the towers.
Hakimi adds that the lack of literacy among voters and diminished confidence in the biometric voting system "can provide ample opportunities for fraud."
Suspicions of fraud have been fueled, in part, by improbable turnout figures in the provinces.
Nationwide, officials say around 2.7 million -- or about 26 percent of eligible voters -- took part in the election. The historic low turnout was seen by some observers as a sign that widescale ballot stuffing had been prevented.
But the provincial breakdown of turnout figures provided by the IEC has fueled suspicions, revealing inflated numbers in some areas -- such as Nangarhar -- and unexpectedly low ones in others.
Nangarhar was above the national average with 30 percent turnout.
In Helmand, the Taliban's stronghold and the most violent province in the country, turnout was 35 percent.
In neighboring Kandahar, another violence-stricken province, turnout was around 35 percent. In the eastern province of Paktia, another unstable area, turnout was 40 percent.
The relatively high turnout in the eastern and southern provinces -- where violence and the presence of militants are higher -- has raised some eyebrows. The region is predominately Pashtun and is Ghani's support base. The Taliban, predominately Pashtun, attempted to disrupt the vote with scores of attacks on election day.
But seemingly inflated figures were also reported in northern provinces that are the major support base of Abdullah. In Baghlan, there was a 43 percent turnout despite recent fighting in the province.
But northern and western regions -- where there is generally less violence and the population is predominately non-Pashtun -- had exceptionally lower turnout compared to previous elections.
Meanwhile, 446 of the 689 polling stations -- around 65 percent -- that were closed on election day were in the country's 12 northern provinces. In comparison, some volatile provinces in the south and east like Helmand, Nangarhar, Kunar, Nimroz, and Paktika had either a single center shut or none at all.
To raise suspicions further, there was a huge jump in some provinces from the turnout figures given on polling day -- and those reported the following day.
The commission, for example, reported a total of 22,813 votes from 309 polling centers in Nangarhar on September 28. A day later, the commission reported 254,871 votes from 390 polling centers -- more than 10 times the number reported on polling day. Meanwhile, there was also an unrealistic spike in Helmand from a turnout of 18,641 votes from 68 polling centers to 70,218 votes from 61 polling centers.
One of the biggest irregularities in the election occurred even before the vote.
The national voter-registration figure of around 9.7 million was widely seen as inflated. Those figures were used in the 2018 parliamentary elections, when observers said it generated ghost votes.
After the vote, the former heads of both the IEC and the IECC were jailed for five years on charges of fraud and misuse of authority. Eight commissioners were also imprisoned.
But the same flawed voter-registration figures were used again in the September election.
"They failed in the proper registration of voters and once again we suffered this time when people were unable to see their names on the voter registration lists," says Haroun Mir, a Kabul-based analyst. "This is a solid indication that the voter registration lists in many places were inaccurate and false."
The provincial voter-registration breakdown has raised questions about the fairness of the election.
"Provinces in the east and south had higher registration compared to similarly populated provinces in the west and north," says Mohib Iqbal, a senior research fellow at Australian National University. "Now, we see that more votes were cast in the east and the south. The very low turnout in some of the northern provinces is very difficult to justify."