We Logged Them as Weather, and I Typed the Entry Myself

I want to say up front that I am not comfortable writing this. I did twenty-one years in the United States Navy, most of it as an Operations Specialist, which for the civilians reading means radar. I sat in dark rooms and watched scopes and talked on radio circuits and kept track of everything flying and floating within a couple hundred miles of the ship. I retired a chief. I loved the Navy. I still love the Navy, and part of loving an institution is knowing what it can absorb and what it can't, and stories like this one are in the second category. So I'm going to be careful. I'm going to tell you what I saw on my equipment, which I knew cold, and what I heard on my circuits, which I can still hear in my sleep, and I am going to keep drawing a line that some of you will find annoying. Here is the line. I can tell you what the scope showed. I can't tell you what was out there. Those are two different things, and twenty-one years of doing the job is exactly what taught me the difference.

In the spring of 2003 I was an OS2, an Operations Specialist Second Class, on a destroyer out of San Diego, in the western Pacific. I'm not going to name the ship. Anybody who was aboard that deployment will know her by the end of the second paragraph anyway, and the ones who weren't don't need it. She was a Spruance-class destroyer, which means nothing to most people now because they're all gone, sunk as targets or cut up for razor blades, but in her day she was a big gray greyhound of a thing, five hundred sixty-some feet, gas turbines, and up in the superstructure a rotating air-search radar that swept the sky out past two hundred miles. That radar is the main character of this story, more than me.

I was twenty-six that spring. I'd been in eight years, made second class on my second try, and I was what they call a track supervisor in the Combat Information Center, CIC, which everybody just calls combat. If you've never seen a CIC, picture a room the size of a doublewide with no windows, blue lights, maybe fifteen consoles with round and square screens, a big status board, the smell of old coffee and electronics and somebody's microwave popcorn from two hours ago. It is the least dramatic room you have ever been in, ninety-nine percent of the time. People do crossword puzzles between reports. That's not a scandal, that's the job; the job is boredom punctuated by radio calls, and you learn to love the boredom because the alternative is somebody shooting at you.

The reason I'm writing this now is that I told the story out loud, all the way through, exactly once, at a funeral in 2018, and the man I told it to had been there, sort of, in a way I'll get to at the end, and what he said back to me has been sitting in my chest for eight years like a swallowed stone. My wife says write it down or let it go. I can't let it go, so here we are.

Some housekeeping first, because the true-or-fake voting on this site apparently gets ugly and I'd rather you argue about the right things. I'm going to change every name, including the exercise name. I'm going to be deliberately fuzzy about dates inside a two-week window, partly because operational stuff from that era is still not mine to give away and partly because, honest to God, I don't remember which night was which anymore. I kept no diary. Personal electronic devices in CIC were prohibited, and I'll tell you later about the one time I broke that rule, which is the single dumbest thing I did in twenty-one years and which, in a beautiful twist, produced no evidence of anything. If you want to call the story fake because a man in his fifties can't line up five specific midwatches from twenty-three years ago, go ahead. The people who remember every detail perfectly are the ones I don't believe.

The deployment itself was ordinary until it wasn't. We left San Diego in January, did the usual transit, Hawaii, crossed the line for the chiefs to have their fun, pulled into a couple of ports I won't list, and by spring we were operating with a carrier battle group in a stretch of the western Pacific that I will only locate for you as deep water, far from any coast, the kind of place where the chart is just blue with tiny numbers in the thousands of fathoms. We were the air-defense picket a lot of the time, stationed out from the carrier, which meant my watch section owned the air picture. That's the phrase we use, owning the picture. Every contact that flew got a track number, a symbol, a history. The system that did this was called NTDS back then, Naval Tactical Data System, ancient even in 2003, symbology like an Atari game, but it worked, and the ships and the aircraft all shared their pictures over radio links so that in theory everybody saw the same thing. In theory. Hold onto that, because this whole story lives in the gap between ships that saw a thing and an aircraft that didn't.

The exercise was a multi-day, multi-unit thing, us and the carrier and a cruiser and a couple other smallboys plus aircraft, and the part that matters for you is that for about a week and a half we were more or less in the same patch of ocean, running the same watch rotations, with the same geometry night after night. Same stations, same radar horizon, same everybody. That's actually rare. Usually you're transiting, the picture changes daily. This was more like a laboratory, which is the only reason a pattern had a chance to show up at all, and believe me, I have thought about that. If we'd been moving, each night's weirdness would have been one sea story apiece, spread across five ships' worth of watchstanders, and nobody would ever have put it together. Instead it was the same dozen of us in the same dark room, five nights out of nine. Or eleven. I've had a guy tell me the exercise was nine days and I remember eleven, and neither of us would bet a paycheck.

I should also tell you what else was going on that week, because when people tell stories like this they clean out all the normal life, and the normal life is what I actually remember best. Our evaporator was broken. The evap is the machine that turns seawater into fresh water, and when it goes down you go on water hours, which means the showers get secured except for certain windows, and when you do shower it's a navy shower, thirty seconds of water, soap up dry, thirty seconds to rinse. By night three of water hours a destroyer full of two hundred eighty people smells exactly like you'd think, and the engineers walked around with a look that dared you to say one word to them. I remember the water hours better than I remember some of the radar stuff. That's just true, and I'm leaving it in because it's true.

Also my wife, who I'll call Renee, was mailing me cassette tapes. This was 2003, we had email on the ship, sort of, a slow trickle that came in batches, but Renee is a talker and not a writer, so she'd sit in the kitchen back in San Diego with a little tape recorder and just talk, forty-five minutes a side, about her sister and the neighbor's dog and the Padres, and mail them out, and they'd catch up to us in these lumpy padded envelopes weeks stale. I had a yellow Sony Walkman, the sports one, and I'd lie in my rack with the curtain shut and listen to my wife narrate a life I was missing. One of the tapes got eaten by the Walkman that spring, chewed the ribbon right out, and I sat on the deck of the berthing compartment winding it back in with a pencil like it was surgery. I mention the tapes now because they come back at the end of this story in a way I'm not proud of.

And there was a card game. Aft berthing ran a continuous spades game that deployment, quarter a point, which was against the rules and everybody knew about it including probably the captain, and about the time the strange stuff started, the spades game blew up into a feud because a seaman I'll call Prue accused my buddy Dagle of talking across the table, coded talk, mentioning cold drinks when he was long in clubs, that kind of thing. This became the dominant social issue on the mess decks for two weeks. Grown men not speaking. I want you to understand that on the same ship, in the same week, we had contacts on the scope doing things that I still cannot explain in 2026, and the thing the crew argued about at dinner was whether Dagle said the phrase ice cold on purpose. Ships are like that. It's the most human thing I know about the Navy and I miss it every day.

Okay. The radar. I promised I'd explain things, so let me build you the picture the way I'd build it for a new seaman, because if you don't understand how the equipment works, the story is just spooky noises, and the whole point is that it isn't.

Our air-search radar was a rotating antenna, up on the mast, sweeping in a circle. Ours went around roughly once every eight seconds. I'm rounding on purpose, and the exact rate was settable anyway, but call it eight. That number is the heartbeat of this entire story, so hold it. Every eight seconds, the beam comes back around to a given patch of sky and paints whatever is there, and the return shows up on my scope as a little slash of video, and the system, or the operator, marks it, and eight seconds later there's a new slash, and the distance between slash one and slash two tells you speed and course. That's it. That's radar. Everything else is refinement.

Two things follow from the eight seconds, and I need you to actually sit with them because internet people always skip this part. First, anything that happens in less than eight seconds is invisible to us. Not blurry. Invisible. The radar isn't a movie camera, it's a strobe light in a dark ballroom; you get one flash of where the dancer is, then darkness, then another flash. Between flashes the dancer could do anything, in principle, but real dancers, real airplanes, obey physics, so the flashes line up into smooth, boring lines, and that smooth boring line is what a track is. Second, when the flashes don't line up into a smooth line, the system and the operator both start from the assumption that the radar is lying, because usually it is. Radars lie constantly. They lie about rain, about flocks of birds, about waves, about each other's transmissions. Half my job, honestly the better half of my skill as an OS, was knowing the difference between a lie and a dancer.

One more piece and then I'll get to the first night. Our air-search set was two-dimensional. It gave range and bearing, meaning how far and which direction, but not height. This surprises people. On TV every blip comes with altitude in a neat little box. Not on our set, not in 2003. Height, for us, came from somewhere else, and in that exercise the somewhere else was the cruiser, who had a three-dimensional air-search radar and shared her picture with us over the link, the radio network that let the ships pool their tracks. So when I tell you, in a minute, that something was at sixty-plus thousand feet, understand what I'm actually claiming. I am claiming that the cruiser's height-finding radar measured that, and reported it on the link, and that my own two-dimensional set held a return on that exact bearing and range at the same time. I never measured the height myself. My gear couldn't. If you want a thread to pull, jury, there's your first one, and I handed it to you myself, because the guys on this site who pretend their gear did things it couldn't do are the reason I almost didn't write this.

The first night, and I believe it was the first night, though it might have been the second because the first one might have been the night the evap quit, I had the midwatch, which runs from about eleven at night to about three in the morning, and the air picture was almost empty. The exercise flying was done for the day. The carrier had a couple of aircraft up doing their own thing, all identified, all friendly symbology, all boring. Commercial air was a handful of tracks way up north on the great-circle routes, airliners crawling along at four-eighty knots at thirty-five thousand feet where they belonged.

At some point in the middle of the watch, my raw scope showed a paint at a range I remember as being about forty miles from us, roughly southwest, where nothing was scheduled and nothing had been. One slash of video. No big deal, at all. Single paints happen every sweep, somewhere. Sea return, a bird with a wet wingspan, a lie. You wait eight seconds. It painted again, same spot, maybe a hair different. Waited again. Third paint. Now that's a contact, so I hooked it and started a track, which took all of five seconds of muscle memory, and I remember being mildly pleased to have something to do.

The system tried to compute a course and speed and gave me garbage. Speed that jumped between basically stationary and several hundred knots, course spinning like a bad compass. That, also, is normal for a new track with junk video. What was not normal was the link. A few sweeps in, the cruiser's picture reached out and correlated with my track, meaning her radar held something at the same spot, and her track came across with height data attached. The readout said, and I am giving you the number I remember and have repeated to myself for twenty-three years, sixty-two thousand feet.

I looked at that the way you'd look at a gas pump that said your tank took ninety gallons. Sixty-two thousand feet is not a place airplanes hang out. Airliners live in the thirties. Fighters can zoom up through the fifties and low sixties for a minute if they're showing off, burning gas like a bonfire, but they don't loiter there. A U-2 lives up there, and we had no reason on earth to think anybody's U-2 was over that patch of empty ocean, and a U-2 does about four hundred knots, nice and smooth. This thing's speed solution was garbage precisely because it wasn't moving smooth. It sat, more or less, wobbling around a point, at sixty-two thousand feet, for I want to say four or five minutes. I called it out to the watch officer, the TAO, tactical action officer, who came over and looked, and said something like, is that real, and I said the forty's painting it and the cruiser's holding it with height, and he said huh, and that huh is the most honest thing anyone said all week.

Then it did the thing. And I want to walk through the thing slowly, because this is the part where I do the math for you, and the math is the whole story.

Sweep N, the paint is there, forty-ish miles out, cruiser height readout sixty-two thousand. Sweep N plus one, eight seconds later, the paint on my scope has stepped inward, toward us, by a little over a mile, and the cruiser's height on the link reads sea level. Zero. Not a lost track, not a dropped return. A return, at the surface, on the same bearing.

Now. On my two-dimensional scope alone, that inward step is odd but not crazy, and I need to be honest about that. My set measures slant range, the straight-line distance from my antenna to the target, like a string stretched through the air. A target forty miles away horizontally and twelve miles straight up is at a slant range of about forty-one and change. If it falls straight down to the water and the horizontal distance doesn't move, the string shortens to forty even. So a vertical drop shows up on my scope as the paint stepping about a mile closer in one sweep, which pencils out to something like five hundred knots inbound if you insist on reading it as horizontal motion, and five hundred knots inbound is fast but it's a Tuesday. If it had only been my scope, I'd have logged a jittery contact and gone back to my crossword, and I want that on the record.

But the height. Sixty-two thousand feet is a little under twelve miles. Twelve miles of vertical in one sweep interval, eight seconds, is a shade over five thousand miles an hour, straight down. That's around Mach 7, plus or minus, depending how you count, pointed at the ocean. And here's the part that I have never once heard a debunker engage with, in twenty-three years of occasionally, cautiously, floating this story around people who work with radar. It's not the speed. Things can go Mach 7; a reentering warhead does. It's the stop. The next sweep after that, and the one after, the return was still there, at the surface, at essentially the same spot. Whatever came down did not hit the water at five thousand miles an hour, because when things hit the water at five thousand miles an hour there is an event. A splash like a building. A datum, wreckage, disturbed sea that the surface-search radar chews on for minutes. The surface set held a small, calm return, like a fishing boat, for about a minute. Then nothing. Sea return. Empty ocean.

Eight seconds. I've done this math on napkins, on the back of a gas receipt in a parking lot in 2018, on my phone calculator at red lights. It doesn't improve. Either the cruiser's height finder told a spectacular lie at the exact moment my independent radar showed a range step consistent with the same geometry, or something went from the stratosphere to the surface of the Pacific in less time than it takes to read this sentence out loud, and arrived gently.

I logged it. I wrote, and I remember the wording because I got asked about the wording later, air contact faded, correlating surface contact faded, at time whatever. Faded. That beautiful Navy word. The TAO stared at the scope for a while and told me to note it in the pass-down, and the watch turned over, and I went down to the mess decks and ate a slider and listened to two guys argue about whether Dagle had said ice cold on purpose, and then I lay in my rack and did not sleep, and that was night one.

The next day I did what you do, which is mention it at chow like it was funny. Hey, anybody else's gear drunk last night? And the day-watch guys had the normal stories, the forty being the forty, and I let it drop. I want to be fair to myself about why I let it drop. Nobody hushed me, and I wasn't scared. Single-night weirdness is just genuinely, boringly common. If you stood a thousand midwatches, and I stood a lot more than that, you have seen the gear do things that would make a civilian call the newspapers, and there's almost always a reason, and the reason is almost always dumb. A tech with a screwdriver halfway through an alignment. Two ships' radars beating against each other. The atmosphere doing atmosphere things, which we're getting to. One night is nothing. I'd have forgotten night one by Guam.

Night two, or the night I'm calling night two, I wasn't even on watch when it started. I had the evening watch and got relieved before midnight, and I was in the berthing passageway in my towel waiting on the water hours window, and Dagle, who had relieved me, sent a seaman down to get me. The seaman, it was Prue actually, of the spades feud, said Dagle says come up, your friend is back. I remember being annoyed, because the shower window was only an hour, and I remember going up anyway in my flip-flops with my towel over my shoulder, and I stood in the blue light of combat in shower shoes like an idiot, which was not the dumbest I looked all deployment but was in the running, and there it was.

Same general area, southwest, different range, closer to fifty miles this time. Same behavior in a general sense. Paint, loiter, height from the cruiser in the high fifties this time, then a drop, though this drop I did not see with my own eyes on the sweep it happened; I got there a couple minutes after, and what I saw was the aftermath, the surface-track phase, the little calm fishing-boat return sitting where nothing should be, and Dagle talking on the coordination net with the cruiser confirming they'd held the same thing. So for night two, the actual drop, I'm telling you what Dagle told me across three feet of air thirty seconds after it happened, with his finger still on the scope. Believe him or don't. He's dead now, which is part of why I'm writing this, and I'd trade a lot to have him here to type his half.

Here's what made night two the night it stopped being funny. The E-2 was up. The E-2 Hawkeye is the Navy's flying radar, the prop plane with the big dish on its back that lives on the carrier, and its whole reason to exist is that a radar at twenty-five thousand feet sees over the curve of the earth in a way surface ships never can. The E-2's picture is supposed to be the best picture. When the E-2 and a ship disagree about a low contact far away, you believe the E-2, every time, because we're peeking over a curve and they're looking down from a balcony. Everyone in the business has this drilled into them from A school forward.

The E-2 held nothing. Not during the loiter at altitude, not during the surface phase. Their controller came up on the net, and I heard this part live, and said in that flat aviator voice, negative contact that posit, and then a beat and then, confirm your track, like maybe we'd fat-fingered the symbology. And the cruiser confirmed and we confirmed, two ships, two independent radars, and the balcony held empty air.

People who know radar are already leaning forward, because that's backwards twice. It's backwards for the surface phase, because a small surface target fifty miles from me is at or past my radar horizon, down behind the curve, exactly the thing I should struggle with and the E-2 should own. And it's backwards worse for the altitude phase, because a target at sixty thousand feet is above everybody's horizon for two hundred miles. There is no geometry, none, in which two ships hold a legitimate high-altitude target and an airborne early warning aircraft on station cannot. When the smart petty officers on my ship, and I flattered myself I was one, tried to build the story where our gear was right and the E-2's was right at the same time, the only way to do it was if the thing on the scope was not where the scopes said it was. Which is a real category. It has a name. And this is where I have to be honest in a way that the true-believers in the comments are going to hate, because the name is anomalous propagation, and it was absolutely, genuinely, no-kidding present that week.

Let me explain ducting like I'd explain it to my grandson, because it deserves a fair hearing. Radar energy travels in more or less straight lines through air that's all the same. But when the atmosphere gets layered, warm dry air sitting on cool moist air, which happens over oceans all the time, especially in the kind of high-pressure weather we were sitting in, the layers can act like a pipe. A duct. Radar energy that should have flown off into space instead gets bent down, trapped, and carried along the surface for absurd distances, and then it hits something real, a ship, an island, a rain squall, two hundred or four hundred miles away, and comes back down the pipe to your antenna. Your radar, honest machine that it is, reports a target at the bearing it heard the echo from, at a range based on the time delay, and it draws you a nice believable blip that corresponds to nothing at that location. Ghosts. Every OS has seen them. In heavy ducting the scope can look like a false armada. The meteorology guys measure the layers with a weather balloon and put out a report, and I pulled the reports that week, and yes, we had ducting conditions on and off the whole time. There it is. That's the jury's Exhibit A against me and I'm the one entering it into evidence, because if I hide it and you find out later, you'll throw out the parts of my story that deserve better.

And I'll go further, because fair is fair. Two of the five nights, I think ducting explains cleanly. Nights, I'll call them three and five, though I told you already I'm shaky on the order, we had surface-ish contacts at strange ranges that faded, no height data, no drops, nothing that a textbook duct doesn't cover, and if those had been the only nights, this story would not exist. I've been the chief myself since then, standing behind some excited kid pointing at a ghost, and I've said the words myself, that's ducting, log it and drink your coffee, and I was right every time I said it. Every time I said it, it was ducting.

But ducting is a surface phenomenon. The pipe hugs the sea. It makes false surface targets and false low-altitude targets. It does not put a steady return at sixty-two thousand feet on a height-finding radar, because the duct isn't up there; the whole mechanism lives in the bottom few hundred feet of air. And it especially does not make a high return become a surface return in one sweep in a way that two radars of two different types on two different ships agree about while a third, better-placed radar sees nothing and then, this is the part, also makes the E-2 wrong in the opposite direction. Ducting explains ships seeing too much. It has no mechanism for the E-2 seeing too little at altitude. I've drawn this on napkins for people. The napkin always ends up with two arrows pointing opposite ways and me saying pick one, you can't have both, and the honest ones go quiet and the other kind change the subject to swamp gas.

Our chief in 2003, I'll call him Chief Manzano, was one of the honest ones, which is why what he did next sat wrong with me for years before I understood it. Morning after night two, he read the logs, took me and Dagle to the little space behind combat where the coffee mess was, and said, in words I'll reconstruct as fairly as I can, we log what explains itself. That's not what he literally said. What he literally said was something like, the met report says ducting, so it's ducting, and when either of you writes anything down it says anomalous propagation, and if it does something ducting doesn't do, you call me, you don't editorialize in a legal log. And Dagle, who had watched the drop live, said Chief, and got one syllable out, and Manzano said, I know what you saw. Log it as AP.

I know what you saw. Not, you didn't see that. Not, knock it off. I know what you saw, log it as AP. I've chewed on that sentence for two decades. At the time it felt like a lie being ordered, and I was twenty-six and righteous about it in the privacy of my skull. Now, having been a chief, having written logs that got subpoenaed in a collision investigation, God forbid you ever live that, I hear it differently. A ship's log is not a diary. It's a legal document with no room for wonder in it. AP was the truthful available category, the way accident is a truthful available category on a death certificate. He wasn't telling us what was out there. He was telling us what the paperwork could carry. I resented him for years and I was wrong to, mostly. Mostly.

Somewhere in the middle of all this, and I genuinely cannot pin the night, the EW module got involved. EW is electronic warfare, and on our ship it was two guys in the corner of combat with a piece of gear everybody called the Slick, which listens. It doesn't transmit, on most days; it sits there with its ears open, and when any radar anywhere paints our ship, the Slick hears the pulse, measures it, and compares it against a library of every emitter the intelligence world has ever cataloged. Every radar has a fingerprint, and one of the main whorls in the fingerprint is something called PRF, pulse repetition frequency, which is how many pulses per second the radar sends. I am now at the edge of my actual knowledge and I want to flag that clearly. I was a scope man, not an EW man. What I know about PRF is what a first class I'll call Kohlmeyer told me across a coffee pot, and I'm passing you secondhand goods, marked as such.

What Kohlmeyer told me, the morning after one of the drop nights, leaning against the bulkhead with his coffee in one of those brown wardroom-reject mugs, was approximately this. During the window when we held the contact, the Slick took hits from an emitter on that bearing. Not continuous, a few intercepts, seconds each. The system went to the library and came back with nothing. No match. And a no-match is interesting but not alarming by itself; new emitters happen, libraries lag. What made Kohlmeyer stand there and drink coffee at me was the closest-match candidate the system offered, which was, and he laughed when he said it, a weather radar. Some commercial precipitation radar, the kind of thing that lives on top of an airport. Right family of frequency, he said, right general everything, except the PRF was wrong, and wrong in a direction he then explained to me twice and I did not fully get either time and I'm not going to fake it for you now. The gist as I retained it was, the pulse rate was way outside what that kind of set uses, so far outside that it changes what the radar could be for. He said something like, it's like hearing a heartbeat that matches a hummingbird, in a chest the size of a cow's. Then he said, or possibly I've added this over the years, whoever built that, I want their catalog. Kohlmeyer if you're reading this and I've mangled your physics, write your own account, I'll vouch for you.