There are two ways this comes up. The first is very insulting, and usually takes the form of: "You poor, sad little man. Atheists don't believe in anything they can't touch, and therefore they can't believe in love or justice or [insert wonderful non-physical thing here]." Of course, atheists are not robotic, emotionally-stunted weirdos who don't believe in anything they haven't seen - no matter how many religious types buy into that (self-comforting) stereotype.
There are plenty of theists who realize this, of course, and some of them decide to insult atheists in a different way, by saying, "You atheists claim to want scientific evidence for God(s), but you don't demand the same kind of evidence for things like love, justice, etc. You're intellectually inconsistent."
This, too, is a common but false stereotype. Atheists don't demand that everything come down to a measurable value, and don't demand utter certainty on every topic. But it's wrong in another way - there can indeed be evidence for things like love, and it's quite different from the kinds of 'evidence' generally advanced for gods.
For one thing, I have direct experience of love - there are people that I love very much, and who love me. I can feel love, directly. Gods, though... not so much. There are people who claim to do so, but they don't really seem to have good evidence for it. (See here and here.) But even if I were autistic, I'd be able to detect love.
First, let's talk about what it means to detect or measure something.
Once upon a time, scientists were looking at fission events in cloud chambers and saw that occasionally, the particles would go off in strange directions, as if the laws of conservation of momentum and energy weren't working right. It was as if an invisible particle had been emitted, and the reaction pushed the visible ones. They theorized that there was a particle being emitted, but it barely interacted with anything else. Thus, the idea of the neutrino was born. Work was done to figure out what properties this particle would have, and eventually a means of detecting them was developed. We still miss all but an infinitesimally tiny fraction of them, but we've got sufficiently precise estimates of how many the sun's emitting that it was actually a problem for a while that we didn't find all we expected. (It turns out neutrinos have different "flavors", and a tiny amount of (unexpected) mass, so they can "oscillate flavors".)
(Even now, we don't directly detect neutrinos; when they have one of their rare reactions with ordinary matter, we detect particular types of radiation being emitted and infer the presence of a neutrino.)
Sharks and platypuses have an electrical sense that we don't have; they can directly feel electrical fields in the water around them. There's some evidence humans may have a weak magnetic sense, but many migratory birds are quite sensitive to magnetic fields. With equipment we can mimic these senses, but not with the natural fluency and detail of the originals.
Consider the difference between a photographer's eye, their camera, and their light meter. All three detect light. The human eye's the most sophisticated of the three, and has the best dynamic range and color response. The camera does better in a few areas than the eye, but overall, if you had the choice you'd rather see with the eye than the camera all the time. And then there's the light meter - it just indicates overall intensity and maybe direction. It doesn't capture any detail at all, it just registers the presence of light and the rough quantity.
Now, imagine a race of aliens that, say, evolved under the ice cap of a moon like Europa. No light at all, so no eyes evolved. No concept of sight; sonar, perhaps, but no vision. Still, they would eventually discover the principle of electromagnetic radiation - perhaps the first clue would be feeling the infrared heat radiating from volcanic vents. Their first instruments for detecting light would almost certainly be as crude as a light meter, like how our Gieger counters detect the nuclear radiation that's invisible to us.
Supposedly 'love' is not objectively detectable and measurable. Most people who make this claim are a bit fuzzy about what exactly they mean by "love". Apparently eros (intense desire, e.g. lust) is not what they are talking about. Most people make a distinction between "philia" (fondness or appreciation) and "agape" (self-sacrificing love, where someone else's well-being becomes important or essential to one's own). For example, I have "philia" toward my friends, and both "agape" and "philia" toward my children - but I have "agape", "philia", and "eros" toward my wife.
Now, let's just focus on "agape". It might be difficult to detect and/or measure, but not impossible. Imagine some Hollywood alien invaders - perhaps insectlike - intelligent but totally lacking "philia" and "agape". Certainly much of human behavior would be puzzling to them, but I think they'd twig to the existence of love eventually. How else to explain the effectiveness of taking hostages?
I can imagine experiments they could perform - not ones that sane, moral humans would ever consider, but still. "How many volts is this guy willing to endure to keep us from eating his children?" You could get a numeric value for that. "How long will this guy keep bailing water to keep his wife from drowning?"
(We can imagine more ethical experiments, too, of course. "How hard will this guy work to make sure his children are fed, housed, and educated? What is this woman willing to give up to make her mother-in-law happy?" etc. People do have ways to detect love in practice - how many songs are there about the difference between saying you love someone, and actually behaving as if you do?)
Does this measure all of love? Of course not. There are ways to measure sexual arousal, but they don't measure all of lust, either. We're still in early days understanding psychology, cognition, emotions, and so forth. But I think we can conclude that love does have an effect on people's behavior - or else it's not really love, is it?
Perhaps this makes my previous analogy a bit more clear. Yes, the crude and brutal "experiments" I proposed don't measure all of love; they're more like a light-meter or Geiger counter, not a 12-megapixel camera. But they are sufficient to demonstrate love's presence or absence. They can't really differentiate between "philia" and "agape" (not unlike the first neutrino detectors, which could only detect one "flavor" of neutrino) but they can show love in the general sense.
Humans have eyes, and only need light-meters to help set exposures for cameras. Humans also have brains with inbuilt systems for understanding their own emotions and that of others (though sometimes that has trouble developing and autism results). With respect to love, we have the same relation to the "loveless" aliens as we would have with sight and the "eyeless" aliens. We can tell (by the effect on actual behavior) when someone loves us or just says they love us. My hypothetical monstrous aliens would have to use cruder, and even more indirect, measures. But they'd be able to see something was going on, affecting behavior in ways they didn't expect. We can't see black holes except for the effects they have on nearby stars, either, but we can tell something's there.
We can't make a neutrino camera and image what's going on in the heart of the sun, much though we'd like to. That doesn't mean we can't detect neutrinos. And we don't understand enough about minds yet to really construct solid theories and fully testable frameworks yet. But yes, we can show that there's love.