For decades, agencies around the United States have been collecting data on mosquitoes. Biologists set traps, dissect captured insects, and identify which species they belong to. They’ve done this for millions of mosquitoes, creating an unprecedented trove of information—easily one of the biggest long-term attempts to monitor any group of animals, if not the very biggest.

The problem, according to Micaela Elvira Martinez from Princeton University and Samuel Rund from the University of Notre Dame, is that this treasure trove of data isn’t all in the same place, and only a small fraction of it is public. The rest is inaccessible, hoarded by local mosquito-control agencies around the country.

Currently, these agencies can use their data to check if their attempts to curtail mosquito populations are working. Are they doing enough to remove stagnant water, for example? Do they need to spray pesticides? But if they shared their findings, Martinez and Rund say that scientists could do much more. They could better understand the ecology of these insects, predict the spread of mosquito-borne diseases like dengue fever or Zika, coordinate control efforts across states and counties, and quickly spot the arrival of new invasive species.

That’s why Martinez and Rund are now calling for the creation of a national database of mosquito records that anyone can access. “There’s a huge amount of taxpayer investment and human effort that goes into setting traps, checking them weekly, dissecting all those mosquitoes under a microscope, and tabulating the data,” says Martinez. “It would be a big bang for our buck to collate all that data and make it available.”

Martinez is a disease modeler—someone who uses real-world data to build simulations that reveal how infections rise, spread, and fall. She typically works with childhood diseases like measles and polio, where researchers are almost spoiled for data. Physicians are legally bound to report any cases, and the Centers for Disease Control and Prevention (CDC) compiles and publishes this information as a weekly report.

The same applies to cases of mosquito-borne diseases like dengue and Zika, but not to populations of the insects themselves. So, during last year’s Zika epidemic, when Martinez wanted to study the Aedes aegypti mosquito that spreads the disease, she had a tough time. “I was really surprised that I couldn’t find data on Aedes aegypti numbers,” she says. Her colleagues explained that scientists use climate variables like temperature and humidity to predict where mosquitoes are going to be abundant. That seemed ludicrous to her, especially since organizations collect information on the actual insects. It’s just that no one ever gathers those figures together.

Together with Rund and a team of undergraduate students, she found that there are more than 1,000 separate agencies in the United States that collect mosquito data—at least one in every county or jurisdiction. Only 152 agencies make their data publicly available in some way. The team collated everything they could find since 2009, and ended up with information about more than 15 million mosquitoes. Imagine what they’d have if all the datasets were open, especially since some go back decades.

A few mosquito-related databases do exist, but none are quite right. ArboNET, which is managed by the CDC and state health departments, mainly stores data about mosquito-borne diseases, and whatever information it has on the insects themselves isn’t precise enough in either time or space to be useful for modeling. MosquitoNET, which was developed by the CDC, does track mosquitoes, but “it’s a completely closed system, and hardly anyone has access to it,” says Rund. The Smithsonian Institution’s VectorMap is better in that it’s accessible, “but it lacks any real-time data from the continental United States,” says Rund. “When I checked a few months ago, it had just one record of Aedes aegypti since 2013.”

“That’s why we came up with this idea of a national surveillance system,” Martinez says. “The U.S. government should make it a requirement for mosquito-control boards to send in their data.”

Some scientists who work on mosquito control apparently disagree, and negative reviews have stopped Martinez and Rund from publishing their ideas in prominent academic journals. (For now, they’ve uploaded a paper describing their vision to the preprint repository bioRxiv.) “Some control boards say: What if people want to sue us because we’re showing that they have mosquito vectors near their homes, or if their house prices go down?” says Martinez. “And one mosquito-control scientist told me that no one should be able to work with mosquito data unless they’ve gone out and trapped mosquitoes themselves.”

Other scientists aren’t convinced that collating the data would be useful. “To predict an outbreak, a national database would be of no use,” says Lyle Petersen, who directs the CDC’s Division of Vector-Borne Diseases. “Mosquito density is very, very local. Knowing what’s going on with mosquitoes in California isn’t going to help you with what’s going on in Illinois. The data from even one part of a city may be totally irrelevant to another.”

Micah Hahn, an epidemiologist at the University of Alaska Anchorage, feels differently. She says that a national mosquito database “would be extremely useful for public health-planning and emergency response”—but creating one would be a logistical nightmare.

“There are hundreds to thousands of independently operating vector-control units—some not more than a guy and truck in rural areas—with varying levels of human resources and computer savvy,” she says. They vary in which traps and baits they use, whether they count males as well as females, and whether they identify insects down to the species or genus. Merging such disparate data would be hard, and there’s no quality-control system in place for checking it.

Roxanne Connelly, an entomologist at the University of Florida, agrees. “As a researcher, I’d love to have access to a nation-wide mosquito surveillance database and have often wished for such a system in Florida,” she says. But after 17 years of experience, she thinks that “trying to standardize the data, for the past or in for future, is going to be almost impossible.”

Martinez agrees that this is the biggest obstacle to creating the database of her dreams—but she doesn’t see it as a deal-breaker. She and Rund argue that many states have already taken steps toward building open databases. Iowa, for example, makes all its mosquito data freely available. “At a minimum we are advocating for stringing these systems together,” they say.

“Data should be made available without having to justify exactly what’s going to be done with it,” Martinez says. “We should put it out there for scientists to start unlocking it. I think there are a ton of biologists who will come up with cool things to do.”