The U.S. Border Patrol has caught a fraction of the border crossers spotted by a sophisticated sensor mounted on unmanned spy aircraft and flown over remote stretches of desert, casting doubts on claims that the area is more secure than ever, according to documents obtained by the Center for Investigative Reporting.

The border crossers were spotted with a new, all-seeing radar system developed for use in the Afghanistan War and patrolling above the U.S.-Mexico border in parts of Arizona since March 2012. The system can reveal every man, woman, and child under its gaze from a height of about 25,000 feet.

Between October and December, records show, the remotely operated aircraft detected 7,333 border crossers during its Arizona missions. Border Patrol agents, however, reported 410 apprehensions during that time, according to an internal agency report. The sensor was credited with providing surveillance that led to 52 arrests and 15,135 pounds of seized marijuana.

Dubbed VADER (Vehicle and Dismount Exploitation Radar) and conjuring images of the Star Wars villain, the sensor can cover a wide swath of land and follow movement as it happens. The system, which is on loan from the U.S. Army, is used to identify roadside bombers in war zones.

Customs and Border Protection officials, who aim to buy two systems for the agency, have touted the system’s effectiveness and testified before Congress that it is changing the Border Patrol’s long-term strategy on securing the border. Yet its unique abilities could shine an uncomfortable light on the agency’s ability to effectively patrol the border.

The radar system is providing the Border Patrol with an important snapshot to judge what it calls “situational awareness”—what’s actually happening at the border. But it has left the agency grappling to measure its own success and define “security.”

Using the system, remote operators can track vehicles and people on foot in real time and distinguish humans from animals. The technology allows the aircraft to fly above bad weather or dust storms that otherwise might ground it, while it sends signals to ground stations that display the human targets as moving dots or black-and-white images.

The internal Customs and Border Protection intelligence report outlines several limitations of the system, including the obvious—it can’t tell the difference between a U.S. citizen and noncitizen. On-the-ground video and other sensors are sometimes needed to confirm these so-called nefarious tracks.

And simply identifying someone crossing the border is just the first step. On the ground, Border Patrol agents often are not available to respond because of rugged terrain or other assignments. As a result, thousands of people have slipped through. At the Border Patrol, they’re known as “gotaways.”

In one week in January, for instance, the sensor detected 355 “dismounts,” or on-foot movement, on the U.S. side of the border in Arizona. Border Patrol agents caught 125 of those, about 35 percent, while an additional 141 people evaded apprehension and 87 more turned back south to Mexico. Two were unaccounted for. The sensor detections led to more than 1,100 pounds of seized drugs.

VADER “has proven to be an extremely effective system in countering threats and supporting the ground commander’s mission in theater,” Boomer Rizzo, a Department of the Army civilian who helps run the radar program, said in an email. “This sensor can track smaller and slower moving targets that traditional radar systems are not able to effectively operate against.”

As for whether the system’s effectiveness has highlighted failures with the nation’s border security, Customs and Border Protection spokesman Michael Friel said the technology is still being tested and its accounting is being refined.

The initial approach used to count who is caught and who evades arrest after VADER detections “was flawed and reflected an incomplete picture of border enforcement,” he said. “There is no silver bullet in border technology.”

Specifically, the agency hadn’t accounted for apprehensions once the unmanned spy plane was no longer patrolling the area or ones made out of its view, he said.

As Congress once again takes up immigration reform, Obama administration officials and others have pointed to the lowest levels of unauthorized border crossers—as measured by Border Patrol apprehensions—and plummeting crime statistics on the U.S. side as proof that their methods are working.

Conservatives have long said that immigration reform cannot come before the border is secure. Immigration-reform supporters, while acknowledging the political need for border security, say the flood of migrants is a symptom rather than a root cause of complex problems now being addressed by Congress.

Amid this debate, unauthorized border traffic has picked up in recent months in some parts of the country. In the Rio Grande Valley sector in South Texas, apprehensions jumped to 97,762 last year, an increase of 65 percent from the previous year, according to internal records.

Some border experts speculate that more immigrants, particularly from Central America, are crossing in recent months because of excitement about an expected immigration-law overhaul, federal budget cuts, and the recent release of detained immigrants by Immigration and Customs Enforcement.

“The border is more secure than ever? Well, that’s a pretty low bar,” said Michael Nicley, who retired in 2007 as the Border Patrol’s sector chief in Tucson, Arizona. “Border Patrol agents would be the first to stick out their chests and say the border is under control. That’s not what they’re saying. Agents I talk to down here say we’re getting hammered.”

Another recent Border Patrol report offers more insights into what VADER detects and how that information passes from one shift of on-duty agents to the next. The March report highlights various sensor detections—from groups of fewer than 10 to more than 100 south of the border. One group of nearly 20 wore booties to disguise its tracks. More than eight hours after VADER spotted them, they were labeled outstanding and passed to the next shift.

Originally designed for war zone

Defense contractor Northrop Grumman, based in the Washington, D.C., area, developed VADER for the U.S. Army to counter roadside bombs and other improvised explosive devices (IEDs) in Afghanistan by detecting enemy combatants as they planted the weapons.

The program was launched in 2006 with sponsorship from the Pentagon’s research arm, known as the Defense Advanced Research Projects Agency (DARPA), to create and test a new radar system within two years. In total, Northrop Grumman has won about $188 million in related contracts, according to a review of contracting data by the Center for Investigative Reporting.

The Army announced in February that it was awarding a sole-source contract to Northrop Grumman for continued support of two VADER systems in Afghanistan and a third in the continental United States through the end of the year. Northrop Grumman referred questions to the U.S. Army, which said information about the domestic use of the system should come from the Department of Homeland Security.

U.S. Customs and Border Protection, meanwhile, has been involved with the development and testing of the system for years.

In 2009, the system was deployed along a 31-mile portion of the Arizona border with Mexico over five days using a Customs and Border Protection unmanned aircraft, according to a 2011 National Research Council report. The demonstration was “a great success” as the system identified suspicious activities four out of five nights, the report states.

Mark Borkowski, a Customs and Border Protection official, testified before a 2011 House panel of lawmakers that the system demonstrated “significant potential” for helping the agency.

Legislators, in turn, have supported the technology with public statements and budget earmarks totaling millions of dollars. Rep. Candice Miller (R-Michigan), who leads a House subcommittee on border and maritime security, said in a June hearing that she was “very encouraged” by VADER.

“This tool is extremely valuable as CBP seeks to identify and detect changing smuggling patterns,” she said.

At the same time, Miller and other lawmakers have questioned the use of apprehension statistics as the most accurate way to measure border security. They have challenged Customs and Border Protection to do a better job of establishing meaningful measurements of success.

“The truth is that we need to refine and strengthen the metrics we use to determine how secure our borders and ports of entry are,” Sen. Tom Carper (D-Delaware), chairman of the Senate Homeland Security and Governmental Affairs Committee, said in a statement.

Nicley, the retired Tucson sector chief, said a key measurement has always been the number of people who evaded the Border Patrol rather than apprehensions. Gotaways are readily discernible to the agency, he said, but they’re not made public readily.

“It’s easy to ascertain how secure the border is,” he said. “Just compare the number who came across the border and the number who were caught, but that’s not what they want to do. Why aren’t they doing it? The only logical explanation is because the numbers won’t be good.”

Border-crossing data

The Government Accountability Office recently reported that the percentage of unauthorized crossers caught by the Border Patrol has increased in six of the Southwest border’s nine sectors.

Some of that increase has been negligible. In the Tucson sector between 2006 and 2011, for instance, border apprehensions rose 2 percentage points. Agents there in 2011 caught 64 percent of crossers, up from 62 percent in 2006.

The Republican-led House Oversight and Government Reform Committee last year sought similar data from the Homeland Security Department as part of an investigation into allegations that officials had released “false and misleading border crossing data” that understated the volume of gotaways, according to a letter sent to Homeland Security Secretary Janet Napolitano.

In a response to the committee’s chairman, Rep. Darrell Issa (R-California), Border Patrol Chief Michael Fisher denied the accusation, writing, “Any suggestion that USBP data collection methods were altered in order to enhance overall border statistics is patently inaccurate.”

Agents say the Border Patrol sometimes labels subjects as “outstanding” for the next shift to search, rather than gotaways, which obfuscates the number of people who evade apprehension, a poor accounting practice that Nicley, the former sector chief, confirmed.

“It’s bad record keeping. It’s not accurate,” he said. “It doesn’t allow the agents to pin them down.”

Art Del Cueto, president of the Tucson chapter of the National Border Patrol Council, the agents union, said agents are being assigned to agencies like the Drug Enforcement Administration or asked to do tasks like southbound traffic checks for money and guns rather than their traditional immigration-enforcement and drug-interdiction roles.

“They say there’s more agents on the line that ever before, but there are not enough agents available to patrol, because we’re not doing Border Patrol work,” he said. “It is very disturbing.”

Agustin Armendariz contributed to this report.

This story was produced by the nonprofit Center for Investigative Reporting. Contact the reporter at aglantz@cironline.org.