Satellite Messengers: Deciphering Reliability

Read any discussion about emergency beacons in the wilderness and you’ll see a common set of claims. People will point out how one device is more reliable than another, list examples of messages that failed to send (or eventually were sent after a long delay), and a few are sure to chime in about how PLBs are the ultimate in reliability.

I have my own preferences, of course, and like to ensure my perceptions are grounded in reasonable fact. There have been a few technological improvements over the years, as well as a few details not immediately apparent at first glance. I can, for example, discuss the advantages of the Iridium network over the GlobalStar design, but is that really why users of one network might see greater reliability?

For this post I’ll refer to satellite messengers such as inReach or SPOT devices as SEND devices. I’ll reserve the Personal Locator Beacon (PLB) label for devices that use the international COSPAS-SARSAT network.

The Common Issue

In the ideal case any device, PLB or SEND, would have a view of the sky clear to the horizon. In practice that’s not the case; a dense forest canopy can obscure the entire sky whilst a mountain face or canyon wall can effectively wall off a significant portion. Additionally, there is the question of what the satellite is able to do with the signal; can it hold onto it until it sees a ground station, or does it simply discard it if no ground station is also in sight? Thus reference is often made to differences in transmission power and frequency of the devices, as well as the type of network. What is more rarely talked about is what the device itself does when the user hits “send”. Let’s start there.

Transmission Duty Cycles

For simplicity we’ll consider the three most common devices: an inReach, a SPOT, and an typical PLB. Transmission of routine messages is limited to the first two, but all three have a distress alerting function. For our purposes the specific model is not particularly relevant. Here are the schedules:

inReachTransmission is attempted immediately. In case of failure, transmission will be re-attempted at the next scheduled interval (default 10 minutes) until success.
SPOTDevice will wait up to 4 minutes for a GPS fix. Without a fix, nothing is transmitted. With a fix, the device will transmit the message 3 times over a 20 minute period.

The behavior changes for distress alerts:

inReachTransmission is attempted every minute for the first 10 minutes. It is then repeated every 10 minutes if movement detected, every 30 minutes otherwise, until the battery is exhausted.
SPOTTransmission is attempted within 1 minute of activation. It is repeated every 5 minutes until the battery is exhausted.1
PLBTransmission is attempted immediately. Transmission continues until the battery is exhausted (24-48 hours).
1 If the non-emergency “Help” button is pressed, the device will wait up to 4 minutes for a GPS fix, then will transmit regardless. The transmission will be repeated every 5 minutes for one hour.

We can see some interesting differences already. With the SPOT it’s much more likely that routine messages will be dropped in areas with poor GPS reception. Additionally, messages will also be dropped if no satellite is “visible” to the device at the time it attempts to transmit. I expect this accounts for most of the perceived differences in reliability. Note that for help and distress messages the device will transmit even without a GPS fix, and continuing for much longer periods. (While a distress alert without coordinates is rather unhelpful, if tracking points have been dropped or a planned itinerary left behind it should be possible to direct searchers to an approximate location.)

Power and Frequency

Here I’ve listed the frequencies used for communication with satellites. (Many PLBs still transmit a homing signal on 121.5 MHz, but the SARSAT satellites no longer listen for this.)

inReach31.7 dBm nominal (~1.5 W)1.6 GHz2
SPOT23.52 dBm EIRP (~0.23 W)1611.25 Mhz – 1618.75 Mhz
PLB5 W406 MHz
2 Round numbers are taken from product documentation and should not be assumed to be exact; refer instead to Iridium documentation for precise uplink frequencies.

All the frequencies are in a reasonable range for satellite communication; below 100 MHz there is the issue of the signal not being able to reliably penetrate the ionosphere and above 3,000 MHz there are issues with getting through foliage and bending around obstacles. Another key factor in penetrating a forest canopy is the slant angle: communication with a satellite directly overhead would see far less attenuation compared to a satellite closer to the horizon. This will be discussed a bit more in the next section.

Satellite Constellations


Iridium has ~76 operational satellites, not counting spares, and provides continuous global coverage. The satellites follow polar orbits (inclination of 86.4°) at al altitude of 781 km. In January 2003 it was also certified for use with the Global Maritime Distress and Safety System (GMDSS). Each satellite maintains contact with the satellites around it creating a mesh network that can be used to pass along messages to relay back to earth.


Globalstar operates around 48 satellites at an altitude of 1414 km. Their orbital planes are inclined at 52°, providing coverage between 70° N and 70° S latitudes. The system uses a “bent-pipe” model where each satellite operates as a repeater. Thus, coverage does not extend to remote locations as it’s not possible for a satellite to be simultaneously in sight of those locations and a ground terminal which can receive the message. Examples include the Southern Ocean, South Pacific, parts of the Indian Ocean and some areas in Central Africa.


The Cospas-Sarsat Programme is an international search and rescue (SAR) initiative. It currently has three satellite components; 9 geostationary satellites (GEOSAR) at an altitude of almost 36,000 km, 5 satellites in low earth orbit (LEOSAR) at around 800 km altitude, and the recently-added MEOSAR layer which was created by adding transponders to the dozens of GPS, Glonass, and Galileo satellites already used for satellite navigation (altitudes from 19,000 to 24,000 km).

The GEOSAR satellites often provide the first detection of a distress signal; because they do not move relative to the earth’s surface they cannot determine a signal’s location on their own, but they are able to relay coordinates if the PLB includes them in the transmission. Most modern PLBs have this capability. Unfortunately they do not provide polar coverage, and may be blocked by local terrain features such as mountains or canyons.

Conversely, because the LEOSAR satellites are moving they can detect the range and bearing to the PLB via Doppler processing. This results in two possible locations, one to either side of the satellite’s path, and subsequent satellite passes are used to rule out the false location and refine the original location. The resolution of a Doppler position is accurate to within 5 km more than 95% of the time. They cover the entire globe twice per day, but for most land areas the time between passes is ~50 minutes.

Finally, the MEOSAR satellites have excellent continuous global coverage and are able to act like a “reverse GPS” system to locate the source of a signal. Eventually this system is expected to replace the LEOSAR system. The expected accuracy is likely be within 100 metres within 10 minutes of the initial transmission.

A 406 MHz beacon’s position is considered confirmed when two passes by different satellites have confirmed the position, or when a Doppler location from a single pass is within 50 km of a position encoded in the beacon’s transmission.

Closing Thoughts

Each system has its unique features and different ways of handling messaging or distress calls. Much of this is not apparent to the user, particularly one who hasn’t read the accompanying documentation. The reliability of distress features in SEND devices is likely better than one might assume based on ordinary messaging reliability. As always, when reception is critical, give devices a clear view of the sky and avoid shadowing them with your body.


  1. inReach SE+ / Explorer+ Owner’s Manual:
  2. SPOT Gen3 User Guide:
  3. C/S G.003 Introduction to the Cospas-Sarsat System:
  4. C/S G.007 Handbook on Distress Alert Messages for Rescue Coordination Centres…:

Electronic Navigation Tips

As NOAA recently announced their decision to sunset raster chart production over the next 5 years, some may find this useful.

Many people are familiar with using apps such as Navionics for basic navigation.  To me, the functionality of such apps is often severely limited, almost more a toy than a serious tool.  That needn’t be the case.

Instead, I’d like to present some options for effective electronic navigation.  The challenge here is finding applications that support them, of which there are few.  My preference is an app called SEAiq; it has an enterprise version with even greater functionality, but we’ll stick with the more affordable recreational version for now.

SEAiq has a few particular tools that are quite useful:  the ability to create electronic bearing lines (EBLs), variable range markers (VRMs), mariner’s notes, and course vectors for both your own ship and for AIS targets.  Some of these can also be combined, for example you can attach a VRM to an AIS target.

Continue reading

Metric Side Effects

A recent thread on Mountain Project caught my interest, when someone asked why US guidebooks give the length of rappels in feet but ropes are sold in metres.  At the time of writing the thread has just reached seven pages of debate on the merits of U.S. customary units vs. the SI units.  While the thread title mentioned “imperial units”, it’s worth noting that the imperial units are actually different from the U.S. units.  This is why the British pint of 20 fluid ounces is larger than the American pint of 16 fluid ounces.  (Of note, those fluid ounces also differ by about 4%.)  Just to kick a little more sand at the U.S. system, I’ll mention the odd bit of trivia that the international foot and the U.S. survey foot are also very slightly different.  (Only some U.S. states use the survey foot.)

But I digress… the argument online rather quickly diverted into a debate about the merits of decimalization, with at least one poster very much attached to inches divided into halves, quarters, and eighths.  While decimalization is not strictly the same as metrication (or “going metric”), it certainly is strongly associated with it!  It occurred to me that a few other concepts also come along with metrication, which is what this post is about.  I suspect many non-metric folk are largely unaware of these side benefits, so this post is for them. Continue reading

Chamonix 2015, Days 1 and 2

In July of 2015 I found myself in Basel on business.  Only 4.5 hours by train from the Chamonix valley, and desirous to escape the heat wave made worse by the Swiss aversion to air conditioning (it’s not strictly illegal, just rather onerous to obtain the necessary permits), I took a few vacation days to fit in some alpine climbing.  By happy coincidence the mountain guide Mark Houston had an opening in his schedule and I was able to obtain his services for the adventure.

Continue reading

Helmets: My Two Cents

Earlier this summer a question on the Great Outdoors StackExchange site led me to look further into the standards for climbing helmets.  The question itself, a basic “can I use my bicycle helmet for climbing”, was fairly simple and has been asked many times in many places.  The development of lightweight foam climbing helmets has also caused some confusion in this area, with even experienced people confused as to where these new designs sit on the spectrum from traditional hard-shell climbing helmets and ultra-ventilated bicycling helmets.

For example, one person wrote: “there are helmets sold as climbing helmets which are basically one-hit-wonders. Those are constructed similar to bike helmets that are meant to crack as they absorb the force of an impact. Once they are so compromised they are pretty much useless. A proper mountaineering helmet would be one built with high impact plastics and other shock absorbing features that allows them to absorb multiple impacts and keep on ticking.

My immediate thought was this writer has rather unrealistic expectations about both types of helmets.  As you’ll see at the end of this post, the relevant climbing (and cycling) helmet standards call for each test helmet to receive impacts in a few different spots (e.g. two on the crown, one on each side, etc.).  Contrary to many expectations, the bicycle helmet standards seem to have just as much emphasis on these “multiple impacts” as do the climbing helmet standards!  Additionally, most “hardshells” currently on the market are actually single-impact hybrid designs.  So, let’s dig into this a bit more…

Continue reading