The strip at Minto looked fine from altitude. I came around low anyway—standard procedure now, even at places I've landed a hundred times. The runway had a wave in it about two-thirds down, maybe six inches of rise where it used to be flat.

I made three passes, working the calculation. The wave was six inches at the deepest point. My safety margin for that approach is eight inches of clearance. I could touch down shorter, before the buckle, but that means less runway for stopping. Or I could accept the wave, trust the gear to handle it. Both options used to be well within parameters. Now they're both at the edge.

I chose shorter touchdown. It worked.

The wave hadn't been there in April. It was July.

I've been flying the Interior for thirty-two years. Learned every gravel bar on the Tanana, every lake that holds through breakup, every village strip from Fairbanks to Galena. That knowledge used to mean something. Now the map keeps changing. Lakes drain overnight through new sinkholes. Runways develop dips between visits. Ground that was solid in May needs a different approach in August.

At Bethel last month, I talked to a pilot who'd been flying the Delta since the seventies. He said Tununak's runway buckled so bad they stopped landing there entirely. Both airlines refused to use it—declared it too dangerous. The village went without regular air service. He'd stopped flying there himself two months before it closed—said he could feel it going but couldn't prove it, couldn't point to anything specific enough to justify refusing the approach. He was right, but being right didn't help the village.

You build a map in your head when you fly. Not the chart kind—the real kind, made from repetition and observation. You know how wind moves through a valley, which gravel bars hold, how a lake looks when it's safe to land. Takes years.

Used to take years. Now it's months. Sometimes less.

I still make the flights. Fairbanks to the villages, the villages to each other. Eighty-two percent of Alaska communities are isolated except by air. The work doesn't stop because the ground's unstable. What stops is the certainty. I used to know where I could land. Now I know where I could land last time.

Every approach requires the same assessment: low pass, check the surface, look for new subsidence, decide if it's safe. The villages still need their deliveries. I still make the decision.

Other pilots are doing the same math. We share information more now—call ahead, ask about conditions, report what we see. A strip that looked good yesterday might have changed overnight. We always talked to each other. Just more urgent now.

I've started keeping notes. Which runways are degrading fastest, which lakes are draining, which approaches need adjustment. Not official—just tracking what I see. Sometimes I look back at entries from five years ago and barely recognize the landscape I was describing.

Permafrost temperatures are halfway to zero. By 2060 they say everything starts moving faster. Feels like we're past that already. The ground's been shifting for years. Just shifting faster.

The strip at Minto held. I unloaded supplies, talked to the village coordinator about the runway condition. She said they'd noticed the buckling, that the elders were arguing about whether to try patching it or accept they'd need a new strip eventually. One of them said eventually was already here. She didn't know if he was right.

I said I'd keep flying in as long as it was safe.

She nodded. We both knew "as long as it's safe" was getting shorter.

I flew back to Fairbanks in good weather. Smooth air, clear visibility. The kind of conditions where flying used to feel like knowledge—like I understood what I was doing and why it worked. Now it just feels like good conditions. While they last.

I've got three more village runs this week. I'll make the same assessments, the same calls about what's safe. Don't know if that's enough. Don't know what else to do.