Silicon Valley (and Montana) radiation levels

Sensor: LND 7121 G-M tube, gamma sensitivity (60Co) 18 cps/mR/hr, time-to-first-count circuit, usable range 1 µR/h - 600 R/h.

Frequently asked questions (well, not really)

Q: What am I looking at?
A: The measurements of the dose rate in the San Francisco Bay Area in California, expressed in microröntgens per hour. The elevation is around 30 ft above sea level. The setup is predominantly sensitive to gamma radiation and X-rays and is mounted in a wooden-frame utility space close to an exterior air vent. The chart is updated every five minutes or so.

(There is a temporary gap in the monthly data prior to Feb 14 as the computer controlling the sensor suffered an SSD failure and had to be rebuilt.)
Q: What's the red "Montana reference" line on the plots?
A: For mildly amusing reasons, I also have access to data from an identical indoor sensor in Montana, at an elevation of about 3,500 feet above sea level. I figured it may be an interesting comparison.
Q: But what does this mean?
A: Depends, but most likely nothing.
Q: Gee, thanks a bunch. What's normal?
A: The usual background levels at the Bay Area location appear to be around 8-9 µR/h. This is due to cosmic radiation, the decay of natural radioisotopes in the soil, and various man-made releases over the past several decades. Modest excursions above this baseline (10% or so) typically have no discernible cause, but sometimes are correlated with solar activity or winds from more radioactive parts of the world.
Q: Okay, what wouldn't be considered normal?
A: Increases past 20 µR/h or so would be highly anomalous for this location, but not particularly bad for you. Ignoring some ongoing scientific controversy, the effects of ionizing radiation appear to be roughly cumulative. The first statistically observable effect is a 0.6% increase in cancer risk with a lifetime exposure to an extra dose of 10 röntgens (or more correctly, 100 millisieverts). At a dose rate of 20 µR/h (~0.2 µSv/h), it would take about 60 years to reach that threshold.
Q: Isn't an extra 0.6% cancer risk a pretty big deal?
A: Your lifetime odds of developing cancer are already around 50%. Similarly to salt consumption risks, that extra 0.6% may be noteworthy across large populations, and thus be of interest to public health - but it represents a negligible hazard on an individual basis.
Q: What readings would be truly dangerous, then?
A: There is no simple answer; for example, radioisotopes that are absorbed by your body from food can be far more dangerous than ambient radiation levels, and the meter doesn't check your diet. But an acute dose of 100 R (1 Sv) can cause mild radiation sickness; and even when received more slowly, it can substantially increase your cancer risk (+6%). Acute exposure to 500 R (5 Sv) is often lethal.
Q: Wait a moment... why are Montana readings quite a bit higher than in California?
A: SF Bay Area is a coastal region, while the other location is up in the mountains - which means that there is less atmosphere to stop cosmic rays. The difference in measurements is quite substantial, around 40%. If that freaks you out, consider that a commercial flight easily exposes you to about 300 µR/h!
Q: If there's ever a nuclear emergency, wouldn't I just want to die quickly anyway?
A: Hey, it's your life, but probably not. Such events are a lot more survivable than portrayed in fiction - and more importantly, the future that awaits the survivors is not necessarily all that bleak. You may want to check out a free book titled "Nuclear War Survival Skills", published back in the good old days of the Cold War. It sounds goofy, but it is a surprisingly interesting read and it goes through some hard science to debunk many of the folksy beliefs perpetuated by Hollywood. And if you are interested in common-sense preparedness strategies for more plausible risks, check out this guide, too.
Q: Why is the raw data so noisy?
A: Radioactive decay is a stochastic process. The number of particles striking the detector is highly variable and needs to be averaged over a longer period of time to get stable numbers and observe more subtle trends.
Q: What's the difference between röntgens (R) and sieverts (Sv), anyway?
A: Röntgens measure exposure, sieverts try to take into account the effects of a received dose on the human body, depending on the exact type of radiation in question. For gamma radiation, the ballpark conversion rate is 100 R = 1 Sv.
Q: What about websites like,, RadNet, etc?
A: They are cool, but many of them are closed platforms or have other issues - such as very limited graphing / trend analysis, ambiguous units (CPM), etc. I figured that it doesn't hurt to do this my way. I originally also included weather patterns in the charts, but then Weather Underground discontinued their public API, so there's that.
Q: I want this. How do you have it set up?
A: The meter is NukAlert-ER - a relatively fancy, wide-range time-to-first-count device based on LND 7121; it is hard to find these days and I had to hand-craft a Linux driver for it. More accessible choices include cheap but narrow-range USB meters, such as Radex One or GMC-320Plus ($100); MCU-based hacks to intercept readings on the HD44780 bus of wide-range Canberra ADM-300 units that are currently abundant on eBay but have no USB port ($200); or DIY circuits with cheap Soviet SBM-20 or STS-5 tubes (sub-$40, but not truly calibrated). The Geiger counter aside, the remaining components used to be a collection of cron jobs to fetch solar X-ray flux readings from NOAA, query the API for weather data, and then to draw all the charts using gnuplot; that said, both of these APIs are now defunct.
Q: On a related note, can you recommend any home decor accessories?
A: Yes, of course. Check out this page for more.
Q: Who are you?
A: You can reach me at By the way, your lucky number is: 20861497.