The ITM Museum
A Python Implementation of
the Longley-Rice Irregular Terrain Model v1.2.2 (aka v1.3)

Mike Markowski
mike.ab3ap@gmail.com


Philip L. Rice
(1922-1997)
Anita Longley
(1908-1986)
Kenneth Norton
(1907-1992)
Albrecht Barsis
(1916-1983)

Graphs below were created using itm.py, ITM v1.2.2. The coverage map is created using rfl, Radio Frequency Link, which uses itm.py as one part of full link budget calculations. Confidence and reliability (in graph titles) are discussed in Section 6 of the 1982 "Guide to the Use of the ITS Irregular Terrain Model in the Area Prediction Mode," linked to below.


Predicted 90W signal on 146.985 MHz up to 80 km range. 15 kHz ENB and residential noise level (-119 dBm).


Path loss confidences for a given reliability


Path loss reliabilities for a given confidence

Background

In 1968, with a computer at your disposal and some environmental information, ITM (Irregular Terrain Model) was the most thorough RF propagation model available. It dwarfed simpler models, straightforward enough that they can be implemented on a slide rule. Examples of path loss slide rules can be seen lower down on this page. The ITM universe is the great-circle plane defined by two antenna locations and earth center. Minimal information is provided to the model for a path loss estimate. The data requiring the most effort to obtain is terrain elevation between antennas. But terrain data is readily available online, so that is usually not an issue.

After over a half century, the Longley-Rice ITM continues to be widely used. If you're like me, however, you know little about the people themselves. I spent some time trying to find out more about them and include it below. While the model is named after Anita Longley and Phil Rice, it is an implementation of theoretical work from them and also from Norton and Barsis. Photos above are in authorship order of their famous two volume work, "Transmission Loss Predictions for Tropospheric Communication Circuits." It is abbreviated by NTIA/ITS as TN101 v1 and v2 (Technical Note 101, Volume 1 and 2), abbreviations I use below. The report, "A Guide to the Use of the ITS Irregular Terrain Model in the Area Prediction Mode" is abbreviated as Area Guide.

Unique Things Here

Here are three categories of ITM items not found elsewhere and hopefully make your visit worthwhile. The "Re-creations" link is especially helpful for those learning to use ITM and for those wanting to know how well this python version duplicates Fortran results.

1. Python ITM and Programs

2. Newly Typeset in LaTeX

Please report typos, no matter how small.

3. Historical Versions of ITM

These old ITM versions are for historical interest only, especially appealing for retro-computing environments like simh and dtCyber. There is something oddly satisfying seeing them come back to life as if no time has passed.

NTIA/ITS and Other Links

Below are links to the original reports and code at the NTIA/ITS web site:

ITM Software


George Hufford (1927-2014)

George Hufford of NTIA wrote the most recent fortran version of ITM. Compute power was limited at the time and the code is heavily focused on efficiency, something not as necessary today with better hardware and compilers/interpreters. If you have ever tried to convert Fortran IV code to a more modern language you have suffered the challenge of GOTOs jumping into the midst of loops and similar pain. George Hufford's code has none of that and was straightforward to convert. His impressive Fortran code was eventually converted to C++ by NTIA, line by line, making only changes required by language syntax.

When I took the ITS Fortran code and converted it to python, I removed most Fortran efficiency concerns and matched equations from the papers as similarly as possible. I also try to fully annotate the code, indicating which equations or report sections are implemented. My initial line by line python conversion resulted in a single threaded program using the original loop constructs. I next converted it to use numpy, avoiding loops by using vector calculations so that all cpu cores are put to use. This both shortens code and looks more like the equations in ITM papers. Spot testing of subroutines shows that a nice side effect is the code is sometimes 2x to 3x faster than the single threaded version. I've made no comparisons with Fortran or C++ versions but fully expect compiled code to be significantly faster than interpreted python. Conversion is easy, testing is time consuming! See recreations of figures from the 1982 Area Mode report for results of some of the testing.

NTIA/ITS tells me they will soon (possibly 2023 [edit: nope, 2023 has come and gone!]) be releasing python wrappers for their latest C++ code. ITM 2027 is in the works and it will be interesting to see what enhancements it brings.

Biographies of ITM Authors

Lilli Segre, Publications Officer at NTIA/ITS, has generously provided portions of the biographies below. Additional information is from an IRE transactions publication. While Lilli Segre says the ITM model was primarily motivated by Anita Longley, biographies below also include authors of the theoretical 1967 TN101 reports published a year before the Longley-Rice software model.

Phil Rice

[From NTIA/ITS]

Phil Rice specialized in tropospheric propagation, mathematical modeling, and satellite and space communication. He authored or co-authored many publications of international interest. Rice was an author of ITS's Tech Note 101: "Transmission loss predictions for tropospheric communication circuits," for which he received the DOC Silver medal, along with Anita Longley, Ken Norton, and Albrecht Barsis. Tech Note 101 described the site general method of the Irregular Terrain Model. But Rice recognized the need to go beyond the first volume of Tech Note 101, and he was instrumental in the publication of Tech Note 101, Vol. 2 which included methods to unify the calculations. Prior to his work with ITS, Rice worked to set up blind-landing systems for the Air Force. He also acted as an advisor and chair of study group V in the CCIR.

[From March 1962 "IRE Transactions on Communications"]

Philip L. Rice (M ’52) was born in Washington, D.C., on December 25, 1922. He attended Lawrence College, Appleton, Wis., in 1941, and received the B.S. degree from the Principia College, Elsah, Ill., in 1948. During World War II, he was commissioned at the Yale University Air Force Communications School, and spent 18 months in Brazil setting up and operating blind-landing systems for aircraft. In 1948 and 1949, he was employed by the firm of Raymond M. Wilmotte, Inc., in Washington, D.C. Since that time, he has been a staff member of the Central Radio Propagation Laboratory of the National Bureau of Standards. He is Chief of the Tropospheric Analysis Section of the Radio Propagation Engineering Division at Boulder, Colo.

Mr. Rice is a member of the Institute of Mathematical Statistics and the Scientific Research Society of America.

Anita Longley

[From NTIA/ITS]

Anita Longley provided a legacy still widely used today, 50 years later. Her main research focus was tropospheric propagation model testing and development. Models she worked on have important applications to propagation over irregular terrain, and she further examined the effects of climate on long-term variability. Longley co-authored Tech Note 101, and described the concept of the "urban factor" in propagation. She expanded and refined the predictions in Tech Note 101 to create the Irregular Terrain Model (ITM or Longley-Rice) computer program in 1968. In 1982 she undertook an extensive revision to the model and it was re-released. Her publications were always backed up by extensive measurements. The validation that Anita Longley and Rita Reasoner did in 1982 became the gold standard for comparing models to measured data. The ITM model has lasting importance in enabling the communications applications we enjoy and rely on every day. This model is one of the top downloads from the ITS website, and updates of her work continue in many ITS modeling and prediction projects today.

[From NTIA/ITS Telenotes 1971 newsletter, Vol 1, No. 3]

Anita Longley, born in Saakatchewan, Canada, received the B.A. degree in Physics from McMaster University, and the M.S. degree in Physiological Chemistry from the University of Minnesota. After serving as a research assistant and teaching at the high school and university levels, she joined the National Bureau of Standards (CRPL) in 1955.

Mrs. Longley’s main field of research is the development and testing of tropospheric propagation models, with applications to communication over irregular terrain, and the effects of climate on long-term variability.

She is a member of U.S. Study Group V for the International Radio Consultative Committee (CCIR). She is also a member of the IEEE, RESA, and Sigma Xi, and was awarded the Department of Commerce Silver Medal for joint authorship of NBS Technical Note 101 on "Transmission Loss Predictions for Tropospheric Communication Circuits."

Mrs. Longley has also written or collaborated in writing more than 20 other research reports, and with P.L. Rice has prepared many of the reports that are included in the published documents of the International Radio Consultative Committee (CCIR). Last year she presented a talk at the conference sponsored by the Advisory Group for Aerospace Research and Development (AGARD) of the North Atlantic Treaty Organization held in Dusseldorf, Germany. She has presented papers at meetings of the International Radio Scientific Union (URSI), a number of lectures for local short-term propagation courses, and numerous reports to sponsoring agencies.

Mrs. Longley's outside interests include travel, bridge, and time spent with her family.

Kenneth Norton

[From NTIA/ITS]

Kenneth A. Norton was a pioneer in propagation studies. He developed a full set of definitions for system transmission, basic transmission, path antenna gain, and related concepts fundamental to all propagation modeling. In 1936 he developed usable equations for engineers from the Somerfield equation while working for the FCC. In 1959 he developed the concept of transmission loss in an article published in the NBS Journal of Research. "System Loss in Radio Wave Propagation," described the characteristics of tropospheric radio propagation and simplified the idea of signal loss in radio. The equations in the article also allowed engineers to easily compare antennas even at different frequencies. Norton was a co-author of one of NBS’s most famous publications, Tech Note 101, "Transmission Loss Predictions for Tropospheric Communication Circuits;" it was an extension of his earlier work on transmission loss. The Norton surface wave was named for him following his work on AM band radio propagation parallel to the Earth’s surface. He retired as the Chief of Propagation Engineering.

[From March 1962 "IRE Transactions on Communications"]

Kenneth A. Norton (A '29-M '38-SM '43-F '43) was born in Rockwell City, Iowa, on February 27, 1907. He received the B.S. degree in physics from the University of Chicago, Ill., in 1928, and continued his studies at Columbia University, New York, N. Y., from 1930 to 1931.

During 1929 he was with the Western Electric Company, Chicago, Ill. From 1929 to 1934 he was in the radio section of the National Bureau of Standards, Washington, D.C. Then, until 1942, he worked in the technical information section of the Federal Communications Commission, Washington, D.C., where he was responsible for a technical study of clear-channel broadcasting and the initial technical studies leading to the allocation of frequencies to television broadcasting. He was Assistant Director of the operational research group and a Consultant on radio propagation in the Office of the Chief Signal Officer, Washington, D.C., from 1942 to 1943 and from 1944 to 1946. He also served as a radio and tactical counter-measures analyst in the operational research section of the Eighth Air Force in England, from 1943 to 1944. Since 1946 he has been in the Central Radio Propagation Laboratory of the National Bureau of Standards where he organized and was Chief of the Frequency Utilization Research Section. At present he is Chief of the Radio Propagation Engineering Division, Boulder, Colo., where he is currently concerned with radio guidance systems for missiles and satellites, long-range radio communications systems involving transmission via satellites, tropospheric scatter and propagation at very low frequencies.

He has been a delegate to several international radio conferences, including the Provisional Frequency Board, Geneva, Switzerland, 1948, and the High-Frequency Broadcasting Conference, Mexico City, 1948. He was Vice Chairman of the United States delegation to the 1950 meetings of Television Study Group 11 of the International Radio Consultative Committee (CCIR) in the United States, France, The Netherlands, and the United Kingdom. He was a United States delegate to the Interim Study Group Meetings of the CCIR, Geneva, 1958, and to the 9th Plenary Assembly of the CCIR, Los Angeles, Calif. He was also a delegate to the 11th and 12th General Assemblies of the International Scientific Radio Union (URSI), The Hague, The Netherlands, 1954, and London, England, 1960, respectively, as well as Chairman of the Local Assembly, Boulder, Colo., 1957.

Mr. Norton was awarded the Stuart Ballantine medal by the Franklin Institute for his work on radio propagation and FM and television frequency allocations. In 1960 he received the IRE Harry Diamond Memorial Award, the highest award offered to a government employee in the field of radio and electronics. He was cited for "contributions to the understanding of radio wave propagation." In 1962 he received the Exceptional Service Award of the U.S. Department of Commerce for "outstanding contributions and leadership in the field of radio propagation research." He is a Fellow of the American Physical Society, AAAS, AIEE, and a member of the Scientific Research Society of America, the American Geophysical Union, the American Mathematical Society, the Institute of Mathematical Statistics, and the American Statistical Association.

A. P. Barsis

[From March 1962 "IRE Transactions on Communications"]

Albrecht P. Barsis (A '49-M '53-SM '61) was born in Vienna, Austria, on July 21, 1916. He graduated from the Realgymnasium in 1935, and attended the Technische Hochschule in Vienna. He came to the United States in 1939, and received the B.E.E. degree from George Washington University, Washington, D.C., in 1948.

From 1940 to 1942, and from 1946 to 1952, he was employed with George C. Davis, Consulting Radio Engineer, Washington, D.C., where he worked on broadcast allocations and directional antenna design. During World War II, he served as a Radio Technician with the psychological warfare branch of the U.S. Army. In 1952, he joined the Boulder Laboratories of the National Bureau of Standards, Boulder, Colo., where he has been engaged in tropospheric propagation research, especially in the analysis and evaluation of tropospheric propagation data and their application to system design.

Mr. Barsis is a member of Sigma Tau, The Scientific Research Society of America, and Commission II of the U.S. National Committee of URSI. He is a registered professional engineer.

Path Loss Sliderules

RF propagation models continue to evolve, providing better predictions of how Mother Nature affects propagation of radio waves. Today's fading models that incorporate multipath and multiple antennas are the result of continued enhancements of older models. From early smooth earth models, more detail was continually added as understanding increased. The advent of computers allowed the development and testing of ever more complex models.

Around 2017 at a Tech Expo at work where companies show off their services and hand out trinkets, I received a communications slide rule made of cardboard, pictured below, with a 1971 copyright. Notice that one of its outputs is path loss. AEL seems to have gone out of business in 1987, so I wish I had asked the vendor where these slide rules came from. Someone somewhere must have found a box of unopened ones from decades ago, because this is like new.

From yet another event I received the following more recent path loss slide rule, this one copyrighted 1993, from EDO which is now part of Harris.

There is something enjoyable about the physical aspects of problem solving on a slide rule, a mechanically simple tool yet does so much. After playing with the cardboard slide rules, I found myself spending $50 on an aluminum slide rule made by Pickett and designed by Collins that does similar. The Collins instructions, scanned here, are copyrighted 1965. This slide rule fills a niche that might no longer exist, designing microwave links with passive repeaters. After reading up on them, I couldn't resist writing passive repeater python code.