1 of 17

A blackbox view of incoherent scatter radar

Bill Rideout/Phil Erickson/K Cariglia

MIT Haystack Observatory

brideout@mit.edu/pje@mit.edu/cariglia@mit.edu

1

2025 ISR Summer School

2 of 17

Outline

  • The nature of ISR measurements
  • Brief discussion of ISR blackbox
  • Where are there existing ISRs?
  • Existing ISRs treated as blackbox
    • Exercise with simulator - existing ISRs

2

  • New ISRs treated as a black box
    • Exercise with simulator - new ISRs Simulate creating a new ISR

Blackbox ISR Video 1

Blackbox ISR Video 2

3 of 17

The nature of ISR measurements

3

Each measurement

has a range

resolution

All ranges are measured at the same time;

but multiple pulses need: time resolution

The beam width at a given altitude

and the range resolution give the

spatial resolution

Typical

ISR

plot

4 of 17

Blackbox ISR Outputs

4

  • Electron density
  • Electron temperature
  • One (or more) ion temperatures
  • One (or more) ion velocities

Parameters

  • Error bar on each parameter
  • Spatial resolution
  • Time resolution
  • Spatial coverage

Quality of measurement

5 of 17

Treat ISR as a blackbox

  • What are the science outputs?
  • What knobs can you turn at the input?
    • For an existing ISR
    • If you got to build a new ISR
  • Try it yourself with two jupyter notebook tools
    • Existing and new ISR simulators

5

6 of 17

Error bars 101: The nature of ISR measurements

This is a probability distribution, not a signal...

Imaging trying to determine if a coin is fair in a dark room...

Both the number of tries and the chance of mistaking head and tails

needs to be taken into account...

7 of 17

Nature of ISR measurements

What determines the error bar on a measurement?

The number of measurements

The measurement S/N

and

  • Flipping a coin in a bright room only one time tells you little (good S/N - small count)
  • Flipping a coin a million times in a completely dark room tells you little (poor S/N, large count)

8 of 17

8

Altitude

Time

Parameter

being

measured

Where would you think good

data is using signal/noise?

Where would you think good

is using error bars?

Error bars combine S/N with

integration time

9 of 17

ISR blackbox inputs

What can an ISR user typically control with an existing ISR?

Pointing direction pattern

Radar mode (pulse length and coding, interpulse

period)

Integration period (sets count statistics)

Time of year and solar activity during measurement

10 of 17

Pointing direction (monostatic)

Single direction gives best time resolution

Multiple directions in local area gives

vector velocities

Measurements can be combined

into scans

Tradeoff: number of pointing directions

versus time resolution

11 of 17

Integration period

For dish antenna with multiple positions, integration periods must be selected beforehand.

For phased array antenna or single position dish antenna, can be chosen after the experiment is run.

Time resolution is inversely related

to number of beams.

12 of 17

ISR modes - single pulse

Pulse length

Shorter uncoded pulses:

Better spatial resolution,

worse S/N

Longer uncoded pulses:

Worse spatial resolution,

better S/N

Interpulse period

Shorter time increases counts/sec, limited by duty cycle of transmitter and need to have previous pulse not returning signal

Parameters to set:

13 of 17

13

Where in the world are/were/will there be ISRs?

Poker Flat

Resolute N+S

Millstone Hill

Arecibo

Jicamarca

Sondrestrom

Svalbard (EISCAT)

Tromsø (EISCAT)

Kharkov

14 of 17

Where to get ISR data

  • Active ISRs fully in Madrigal
    • Poker Flat (Alaska), RISR North and RISR South (northern Canada)
    • Millstone Hill (Massachusetts), Jicamarca (Peru)
    • Eiscat (Svalbard, mainland ISRs)
    • Arecibo (Puerto Rico) - future uncertain
  • Active ISRs partially in Madrigal
    • Kharkov (Ukraine)
  • Historical ISRs in Madrigal
    • Sondrestrom (Greenland), Malvern (UK), St. Santin (France)
  • Radars that run occasionally in ISR mode
    • Altair (Kwajalein), Shigaraki MU (Japan) - some in Madrigal
    • EAR (Indonesia), PANSY (Antartica)
  • ISR Radars under construction
    • Eiscat 3D (Scandinavia), Sanya (China)

14

15 of 17

Existing ISR simulator exercise

15

https://tinyurl.com/2025ISR

Blackbox ISR exercises

Additional inputs

16 of 17

ISR blackbox inputs for a new radar

What design decisions affect a new monostatic ISR?

  • Radar frequency
  • Aperture (m^2)
  • Peak power
  • Location
  • Steering method and range

17 of 17

New ISR simulator exercise

17

https://tinyurl.com/2025ISR

Blackbox ISR exercises

Ionosphere generated by

IRI model (quiet day)

Full code available

(~650 lines of python)

All equations to be covered in rest of this course