The neighborhood around here is a bowl. All directions are uphill from the lowest point, that being directly in front of the house, the house being located in the center of a downhill slope from back to front. The indoor antenna is a vertical wire 90-degrees perpendicular to gravity level. The ground wave from the antenna produces a signal much like a body of water, meaning that it is strong downhill, filling the bowl, but fades out on an uphill climb. Tilting the antenna so it's 90-degrees perpendicular to the hill's plane improves the uphill signal in the back yard, but the hill across the street rises at an opposite angle and loses some of its signal. A conical antenna with a wide base and narrow tip-top might work in a bowl, but the FCC inspector might fuss about it. This finding regarding hills probably says something about why to put antennas high off the ground. Bottom line: ground waves don't seem to climb hills.
How does the antenna work outside?
An advantage of being in a bowl with hills all around is that there's less background noise. Those hills, which would probably be called levees if there was water on the other side, block out a noticeable amount of AM background signals. This is evident on one of our Part 15 frequencies, 1550 KHz, when we ride the car up and over the hill we suddenly get a station near Branson, MO. Down in the bowl this signal is only heard during the "critical hours," when the sun is first rising or setting at dusk.
Hi John.
In my blog "Indoor/Outdoor Coverage from 1-Antenna" the windowframe antenna is 90-degrees perpendicular to the gravity plane and it puts a stronger signal into the "bowl" than the same type of (vertical) antenna located indoors, but riding the car the same fade-out happens a short distance uphill.