Guest post by PrepperDoc
Many preppers’ post-disaster communications plans are built upon low power (“QRP,” typically 1-5 watts output power) ham radio equipment, able to easily obtain power from small battery or low-power solar sources. They may believe that after a disaster, interference from higher-powered stations, noisy power lines, electric motors, and a host of computers will be squashed, and their 5-watt level signals will easily make all the necessary communications. Depending on their communications requirements, they may be badly disappointed in the real event!
Survivors may have widely varying communications needs which might be broken down roughly into three categories: 1) ability to listen to (and possibly contribute to) news reports to/from undamaged states or nations; 2) ability to obtain same-city, intra-state and next-state-over reports of situation-on-the-ground (5-300 miles); 3) short-range communications within a neighborhood. #1 is easily handled by low-power ham gear (or even shortwave radio receivers) because there may be multiple possible transmitting stations from which to choose; simply find one you can hear. #3 can be handled by direct (simplex) communications using low-power walkie-talkie FRS/GMRS or ham transceivers. (Store several in a Faraday cage!) It is middle-distance #2 — reliable communication/information gathering from 5-300 miles — that is problematic. You may find a network of several reliable early-warning sites in nearby cities and just across the state border, and have a need to maintain RELIABLE (not hit-or-miss) communications with them daily for updates on security issues. It is nice to know of oncoming trouble farther than “smoke-distance.”
VHF/UHF walkie-talkies simply can’t fill this need with their line-of-sight propagation. And ground-wave (limited at any frequency above 3.5 MHz) transmissions will not cover the distance. One report found 7MHz ground wave unreliable even at 15 km. This 5-300 mile range is the realm where Near Vertical Incidence Skywave communications (NVIS), bouncing near-straight-up radio waves off the ionosphere miles above us (usually the F layer but sometimes the E layer) is the only suitable propagation system. 
The properties of the F layer are important to your success. First, it is at least 150 km above the earth, so your signal is going to travel 300 km just to get to the other side of your town. Modeling your antenna as a point-source, your signal is going to be significantly dispersed and therefore much weaker after traversing that 300 km round-trip distance!
Secondly, the F layer has variable ionization (more during the day, and during maxima of the 11-year sunspot cycle) and is only able to reflect signals at any given moment up to a certain “cutoff frequency” that depends on the both the ionization and the angle of incidence. Vertical signals (needed to get to the other side of your city) are the hardest to refract/reflect. The maximum frequency that successfully reflects vertically is called the Critical Frequency. Somewhat higher frequencies may refract at lesser angles — but constrained by geometry, they will come back down much farther away, leaving you with a “skip zone” of impossible communications.
And unfortunately, you probably can’t use the exact OPTIMAL frequency at any given circumstance. Your prospective counterparties are mostly other amateur radio operators. Ham radio equipment typically is designed to work only in certain designated frequency bands — the 3.5-4 MHz (“80 meter”) and 7-7.3 MHz (“40 meter”) are usually the key ones for reliable NVIS communications. During nighttime around sunspot minima, only the 3.5-4 MHz band may be functioning for NVIS. During the day, both 80 and 40 meters may reflect in more years of the sunspot cycle — but now add in the problem that the lower level D layer, activated by sunlight-accompanying xrays, will all but wipe out 80 meter communications. The D layer’s power-absorption declines by the square of the frequency. As a result, you prefer to use the very highest frequency that works, optimally just below the critical frequency. During daytime, the critical frequency may be 7MHz or even much higher, but many ham transceivers off only 7MHz, 14MHz, 21MHz & 28MHz choices. Thus you may have additional D-layer absorption due to sub-optimal communications frequency. During the night, the sun’s xrays disappear, and the D layer dissipates, so 3.5-4 MHz signals, which are usually safely below the critical frequency even during sunspot minima, become much more useful & important. For reliable nighttime NVIS, you probably need 80 meter capability, which requires large antennas for good efficiency (or else higher power). 
The ultimate goal is simply to provide a signal to the desired receiving station that significantly overpowers the NOISE that the recipient encounters. Inexperienced operators may require signal-to-noise ratios of 10 dB or more for successful communications.
Even after an EMP-type disaster, there may be more radio noise than optimistic low-power proponents expected. Why? Because much of the radio noise in the high frequency bands is the result of tens of lightning strikes every second, all over globe, whose radio-signature is carried around the world by the ionosphere just like any other radio signal.  Even after an EMP, this noise source will still exist. Further, ham radio stations in undamaged nations will still be on the air — and likely far busier than ever before! There will be plenty of strong signals with which to contend. Finally, while power lines may be silent and most computers dark, a new source of man-made radio interference may burst forth–dozens to thousands of power inverters of all types providing power to persons all over your city. Even the sine-wave inverters have powerful switching signals as part of their makeup, and I was surprised to find a very troublesome amount of interference coming from my very own backup power system, wiping out weak-signal reception! The addition of a heavy-duty filtering device in my inverter’s power lines to my home knocked this down considerably, but few survivors are going to have prepared this well, so houses all around you may be radiating radio hash. (Consider a device similar to: http://www.amazon.com/Power-Single-Phase-Filter-CW3-20A-S/dp/B00D0U83D8/ref=sr_1_8?ie=UTF8&qid=1428329468&sr=8-8&keywords=power+line+filter )
For NVIS communications, your antenna can also squelch your effort to overcome the noise at your intended recipient’s site: vertical or whip antennas put relatively little power straight up, further damaging your low-powered transmitter’s chances. A very comprehensive investigation in the Netherlands demonstrated that a horizontal resonant dipole at 0.15 – 0.2 wavelengths height was optimal  (corroborating work done in the rainforests of Thailand). For an 80-meter antenna, that means a height in the range of 40 feet; for 40 meters, 20 feet. Survivors with antennas at first-story roof-level may face a significant power loss of as much as 90% of their effective signal (10 dB). Likewise, too-high an antenna (from a skyscraper) may also lower vertically incident power.
You can reduce the effective noise (and thus improve your chances) by eschewing voice communications and moving to narrow-band techniques such as Morse code or digital communications — IF your receiver has the ability to filter more narrowly, your operator has the required experience, and in the case of digital, your conversion equipment survived the disaster. In our group, we have some new operators who simply cannot use these more-powerful techniques, so we are limited to voice (single side band, about 2 kHz bandwidth).
Beginning to see why QRP low-power ham radio may not meet your security communication needs post-disaster? Basically, there were very good reasons why the most popular ham radio gear of the 1960’s and 1970’s offered a full 100 watts of output! Furthermore, what if there is a second EMP strike? Or third? Will your transistorized low-power ham radio is connected to an antenna during one of those strikes because you depend on it for communications? It may well be destroyed. The most impervious gear to simulated EMP attack in testing was vacuum tube gear: the type of transceivers that had the 100-watt output.
So what is documented about successful and reliable short-to-mid-range NVIS communications in the real world? Working in the rainforests of Thailand, with relatively optimized antennas, 15-watt output transmitters were reliable for NVIS communications 80% of the time. My own group found that with newbie operators and horizontal dipoles at various heights, cross-city (30 mile) communications were sometimes possible on voice, and even more likely on Morse code, but that experience made a very big difference. A Netherlands group did extensive research at a near-optimum frequency of 5.39 MHz for their conditions, using a high-power 850-watt output transmitter and had excellent signal to noise ratios of 50 dB in NVIS communications. Their powerful transmitter even showed evidence of a readable signal that may have been carried the other way–traversing almost the entire globe to reach their recipient; but this signal was some 40 dB weaker. Their advantages over many low-power stations were significant: Their 850 watt station was 22 dB stronger than a 5-watt QRP station, had an optimized antenna (possibly 10-20 dB better than a poorer antenna) and optimized frequency (excessive D-layer absorption due to lower frequency might have added another 10-20 dB of loss). Hence their 50 dB signal to noise ratio could easily have been obliterated by a ham operating a 5-watt station (-22 dB), with a suboptimal antenna (-15 dB) and suboptimal frequency (-15 dB) (total degradation: 52 dB) even before considering the difficulties of inexperienced operators. An excellent advisory on NVIS emergency communicates notes success with 25 watt (output) signals. 
My conclusion is that your communications preparations should definitely include a simple wire dipole antenna at 30-40 feet, either resonant or long-wire horizontal dipoles (with antenna tuner) for both 80- and 40- meter ham radio bands, and possibly additional higher frequency bands for daytime use. You should also develop a healthy dose of experience (Morse code ability and a narrow receiver filter would be great!). But it is obviously easier to “turn down” the transmitter power on a 100-watt (or higher) tube type rugged EMP-resistant vacuum tube transmitter to save energy, than it is to try and make a low power 5-watt QRP transistorized transmitter communicate amidst stronger signals and broadband inverter-generated hash interference, while worrying that your equipment might at any time be destroyed by a follow-up EMP strike. So it might be worth it to plan ahead to provide both ham radio equipment and electrical power for a higher power transmitter, even if you do succeed at times with a QRP transceiver.
 NVIS Army FM 24-18. Appendix M with Graphics. http://kv5r.com/ham-radio/nvis-army-fm-24-18/ (An excellent tutorial.)
 HF Near Vertical Incidence Skywave (NVIS) Frequency Band Selection. Accessed at: http://www.idahoares.info/tutorial_hf_nvis_band_selection.shtml
 Bianchi C, Meloni A: Terrestrial Natural and Man-Made Electromagnetic Noise. Accessed at: http://www.progettomem.it/doc/MEM_Noise.pdf
 Witvliet BA et al, Near Vertical Incidence Skywave Propagation: Elevation Angles and Optimum Antenna Height for Horizontal Dipole Antennas. Accessed at: http://www.agentschaptelecom.nl/sites/default/files/2015_-_witvliet_-_nvis_elevation_angles_and_antenna_height_-_ieee_ant_prop_mag.pdf
 Idaho Amateur Radio Emergency Service, HF near Vertical Incidence Skywave (NVIS) Frequency Band Selection. Accessed at: http://www.idahoares.info/tutorial_hf_nvis_band_selection.shtml
Start now to make sure you are staying prepared.