Skip to content

Polar-Orbiting and Geostationary Satellites

CSIR NET Earth Sciences Part C satellite meteorology: Kepler's law, geostationary and sun-synchronous polar orbits, coverage, resolution, parallax and missions.

  • 19 min read
  • Updated
On this page · 14 sections
  1. Prerequisites
  2. The Hook: One Cyclone, Two Views
  3. Orbital Mechanics in Five Lines
  4. The Geostationary Orbit
  5. The Polar-Orbiting (Low Earth) Orbit
  6. Choosing the Right Orbit
  7. Viewing Geometry: Footprint and Parallax
  8. Memory Tricks
  9. Don’t Confuse
  10. Common Mistakes
  11. Practice
  12. Quick Revision Table
  13. Key Takeaways
  14. What’s Next

Prerequisites

The Hook: One Cyclone, Two Views

Orbital Mechanics in Five Lines

For a satellite of negligible mass in a circular orbit of radius \( a \) (measured from Earth’s centre), gravity supplies the centripetal force:

$$\frac{GM m}{a^2} = \frac{m v^2}{a} \ \Rightarrow\ v = \sqrt{\frac{\mu}{a}}, \qquad \mu = GM = 3.986\times 10^{14}\ \text{m}^3\,\text{s}^{-2}$$

The period is the circumference divided by the speed, which is Kepler’s third law:

$$\boxed{T = 2\pi\sqrt{\frac{a^3}{\mu}}}\qquad\Longleftrightarrow\qquad a = \left(\frac{\mu T^2}{4\pi^2}\right)^{1/3}$$

Altitude \( h = a - R_E \), with the equatorial radius \( R_E = 6378 \) km. Three consequences follow at once:

  1. Higher orbits are slower and have longer periods (\( T \propto a^{3/2} \), \( v \propto a^{-1/2} \)).
  2. The orbit is fixed in inertial space (apart from perturbations); the Earth turns underneath it. That is why the ground track of a low satellite shifts westward on each pass.
  3. The mass of the satellite does not matter. A distractor that says “a heavier satellite needs a higher orbit” is wrong.
Altitude \( h \)Radius \( a \)PeriodSpeedTypical use
850 km7,228 kmabout 102 minabout 7.4 km/spolar-orbiting weather satellites
about 20,200 kmabout 26,600 kmabout 12 habout 3.9 km/sGPS navigation
35,786 km42,164 kmone sidereal dayabout 3.07 km/sgeostationary weather satellites

The Geostationary Orbit

Four conditions

A satellite appears fixed in the sky only if its orbit is:

ConditionWhy it is needed
Circularconstant angular speed, so no east–west drift within the day
Equatorial (inclination 0°)otherwise the satellite traces a figure of eight north and south of the equator
Prograde (eastward, with Earth’s rotation)it must turn the same way as the Earth
Period = one sidereal day (23 h 56 min 4 s = 86,164 s)it must keep pace with Earth’s rotation relative to the stars, not relative to the Sun

Putting \( T = 86{,}164 \) s into Kepler’s law:

$$a = \left(\frac{3.986\times 10^{14}\times 86164^2}{4\pi^2}\right)^{1/3} \approx 42{,}164\ \text{km}, \qquad \boxed{h \approx 35{,}786\ \text{km}}$$

which is about 6.6 Earth radii from the centre, or 5.6 Earth radii above the surface. An orbit with a one-sidereal-day period but non-zero inclination is geosynchronous without being geostationary.

What a geostationary satellite sees

From 42,164 km the Earth subtends a half-angle of \( \sin^{-1}(6371/42164) \approx 8.7^\circ \), so the full disc fits in a field about 17.4° wide. The farthest point visible on the surface lies at a great-circle (central) angle of \( \cos^{-1}(R/a) \approx 81.3^\circ \) from the sub-satellite point, but the view there is useless: the line of sight grazes the limb.

The satellite zenith angle (the angle between the local vertical and the line to the satellite) grows faster than the central angle \( \gamma \):

$$\theta = \gamma + \eta, \qquad \tan\eta = \frac{\sin\gamma}{a/R - \cos\gamma}$$

where \( \eta \) is the scan angle at the satellite.

Central angle from sub-satellite pointSatellite zenith angle
30°about 35°
45°about 52°
60°about 68°
70°about 79°

In practice quantitative products are restricted to zenith angles below roughly 60–70°, which means latitudes poleward of about 60° are poorly served even directly north of the satellite. Geostationary satellites cannot see the polar regions usefully. That is the first reason the world needs polar orbiters.

Strengths and costs

StrengthCost
Continuous view of the same disc: full-disc images every 10–30 minutes, sectors every few minutes in rapid-scan modedistance: for the same optics, resolution is roughly 40 times coarser than from 850 km
Ideal for tracking cloud motion (atmospheric motion vectors), the diurnal cycle, thunderstorm growth, cyclone tracks, fog dissipationpoor or no coverage poleward of about 60–70°
Same viewing angle for a given place, so time series are consistentmicrowave imaging and sounding are very hard at that distance (antenna size), so geostationary payloads are mainly visible and infrared
One satellite covers about a third of the globelarge and growing pixel size and parallax toward the edge of the disc

Two operational nuisances recur in statements. Around the equinoxes the satellite passes through Earth’s shadow near local midnight at the sub-satellite point, for up to about 70 minutes a night, and runs on batteries. Around the same seasons the Sun can also appear close to the line of sight, and imagers may be protected from sun intrusion by closing down channels for a short time.

The geostationary ring

About five or six geostationary meteorological satellites, spaced round the equator, give near-global coverage between about 60° N and 60° S:

OperatorSeriesNotes
USA (NOAA)GOES (East and West)the Americas and the eastern and central Pacific
Europe (EUMETSAT)MeteosatEurope and Africa; a second satellite has been kept over the Indian Ocean
India (IMD and ISRO)INSAT-3D, INSAT-3DR, INSAT-3DSIndian Ocean; six-channel imager plus a 19-channel sounder
Japan (JMA)Himawariwestern Pacific, East Asia, Australia
China (CMA)Fengyun-2, Fengyun-4East and South Asia
Russia, KoreaElektro-L, GEO-KOMPSATregional

India’s meteorological geostationary series runs through the multipurpose INSAT-1, -2 and -3 satellites, the dedicated Kalpana-1 (launched in 2002 as METSAT-1, carrying a three-channel Very High Resolution Radiometer), and INSAT-3D (2013), which added the sounder. Data are archived and served by ISRO’s MOSDAC (Meteorological and Oceanographic Satellite Data Archival Centre) and used operationally by IMD.

The Polar-Orbiting (Low Earth) Orbit

Geometry

Polar-orbiting meteorological satellites fly at about 700–900 km, with periods of about 100–102 minutes, so they complete about 14 orbits a day. The inclination is close to 90°, so each orbit passes near both poles; with a cross-track swath of a few thousand kilometres, the satellite sees every part of the Earth at least twice a day (once on the ascending, once on the descending pass), and the polar regions on many consecutive orbits.

During one orbit of period \( T \), the Earth turns through \( 360^\circ \times T/86{,}164\ \text{s} \), about 25.4° of longitude for \( T = 102 \) min. Successive equator crossings are therefore about 25° apart, roughly 2,800 km at the equator. A swath wider than that, such as the 2,900 km or so of the AVHRR imager, leaves no gaps at the equator in a day of ascending passes; a narrower swath leaves diamond-shaped gaps that are filled on the following days.

Sun-synchronous orbits

Earth’s equatorial bulge (the \( J_2 \) term of its gravity field) exerts a torque on an inclined orbit, and the orbital plane precesses about Earth’s axis. The rate of change of the right ascension of the ascending node is

$$\boxed{\dot\Omega = -\frac{3}{2}\,J_2\left(\frac{R_E}{a}\right)^2 n\cos i}, \qquad n = \frac{2\pi}{T}, \quad J_2 \approx 1.083\times 10^{-3}$$

Choose \( a \) and \( i \) so that the plane turns eastward at exactly the rate at which Earth goes round the Sun, \( 360^\circ/365.24\ \text{d} \approx 0.986^\circ \) per day. Then the angle between the orbital plane and the Sun–Earth line stays constant all year. This is a sun-synchronous orbit.

  • For eastward precession \( \dot\Omega > 0 \), so \( \cos i < 0 \): the inclination exceeds 90° and the orbit is retrograde.
  • At 800–870 km, the required inclination is about 98.6°–98.9°.
  • The satellite crosses the equator at the same local solar time on every orbit and every day (for example, northbound at about 13:30 and southbound at about 01:30). The two crossings are 12 hours apart in local time.
BenefitConsequence
Same illumination at each place, every dayvisible-channel images and products are comparable from day to day
Fixed sampling timea single satellite samples the diurnal cycle at only two local times; operational agencies fly morning and afternoon orbits so that at least four times are covered
Near-polar coveragepolar regions are covered on almost every orbit

Strengths and costs

StrengthCost
Close to Earth: resolution of about 1 km or better in imageryeach place seen only about twice a day per satellite; no continuous watch
Global coverage including the polescloud tracking is limited to high latitudes, where consecutive passes overlap
Room for microwave imagers and sounders, scatterometers, radar and lidarimages arrive as swaths that must be mapped and stitched
The backbone of global numerical weather prediction through sounder radiancesorbit drift in old satellites slowly changed equator-crossing time, a known problem in long climate records

Major polar programmes

ProgrammeKey instruments (meteorological)
NOAA POES (TIROS-N series)AVHRR imager; HIRS infrared sounder; AMSU microwave sounders
EUMETSAT MetOpAVHRR, IASI hyperspectral infrared sounder, AMSU and MHS, ASCAT scatterometer
Suomi NPP and NOAA JPSSVIIRS imager, CrIS infrared sounder, ATMS microwave sounder
China Fengyun-3imagers and sounders in the same class
IndiaOceansat-2 (scatterometer OSCAT), SCATSAT-1 (scatterometer) and Megha-Tropiques (see below)

The first weather satellite, TIROS-1, was launched by the USA on 1 April 1960 and returned television pictures of cloud.

Low-inclination orbits: a third choice

Some research satellites deliberately avoid both choices. A low-inclination, non-sun-synchronous orbit drifts through all local times over a few weeks, so it samples the diurnal cycle of tropical rain and convection, and spends more time over the tropics.

SatelliteInclinationPurpose
TRMM (Tropical Rainfall Measuring Mission, USA–Japan)35°first spaceborne precipitation radar
GPM Core Observatory65°dual-frequency precipitation radar, reference for a constellation
Megha-Tropiques (ISRO–CNES)20°tropical water cycle and energy budget; microwave humidity sounder and radiation budget instruments

Choosing the Right Orbit

graph TD
    A["What does the task need?"] --> B{"Continuous watch
of one region?"} B -->|Yes| C["Geostationary
(cloud motion, diurnal cycle,
nowcasting, cyclone tracking)"] B -->|No| D{"Polar coverage, fine resolution,
microwave or active sensors?"} D -->|Yes| E["Sun-synchronous polar orbit
(soundings for NWP, SST,
scatterometer winds, sea ice)"] D -->|No| F{"Tropical diurnal
sampling?"} F -->|Yes| G["Low-inclination precessing orbit
(TRMM, GPM, Megha-Tropiques)"]
FeatureGeostationarySun-synchronous polar
Altitude35,786 kmabout 700–900 km
Period1 sidereal dayabout 100 min
Inclination0°about 98°–99° (retrograde)
Coverageone disc, about 60° N to 60° S usablewhole globe, twice a day
Repeat over a point10–30 min (minutes in rapid scan)about 12 h per satellite
Imagery resolutionabout 0.5–1 km (visible) and 2–4 km (infrared) at the sub-satellite point on current imagersabout 0.4–1 km
Microwave and active sensorsessentially noneyes
Best forcloud motion, nowcasting, diurnal cyclesoundings, SST, global products, polar regions

Viewing Geometry: Footprint and Parallax

Pixel size

The instantaneous field of view (IFOV) is the angle subtended by one detector element. At nadir the ground footprint is

$$\boxed{d = H\times \text{IFOV}}$$

with \( H \) the altitude and IFOV in radians. An imager with an IFOV of 1.3 mrad at 850 km gives a 1.1 km pixel; to get a 4 km pixel from 35,786 km, the IFOV must be about 112 μrad. Away from nadir the pixel grows: on a flat Earth the dimension along the scan grows as \( \sec^2\theta \) and the dimension across it as \( \sec\theta \); Earth’s curvature makes it worse. Pixels near the edge of a geostationary disc can be several times their nadir size.

Parallax

A satellite sees a cloud top in the direction of its line of sight, not vertically above the ground point beneath the cloud. A cloud top at height \( h \), viewed at satellite zenith angle \( \theta \), is mapped away from the sub-satellite point by

$$\boxed{\Delta x = h\tan\theta}$$

A 12 km cumulonimbus top viewed at \( \theta = 50^\circ \) is displaced by about 14 km. For India seen from a geostationary satellite on the equator, tall storms are placed slightly poleward of their true position. Parallax matters when satellite cloud is compared with radar or rain gauges, and when lightning or rain is matched to cloud tops.

Limb effects

At large zenith angles the path through the atmosphere is long. Infrared window brightness temperatures are colder (more absorption by water vapour, emission from higher levels), an effect called limb cooling; retrieval schemes include a \( \sec\theta \) correction for it. This returns on the retrieval page.

Memory Tricks

Don’t Confuse

PairDistinction
Geostationary vs geosynchronousperiod of one sidereal day and circular equatorial prograde orbit vs period of one sidereal day only (may be inclined or elliptical)
Sidereal day vs solar day86,164 s (rotation relative to stars) vs 86,400 s (relative to the Sun)
Polar vs sun-synchronousinclination near 90° vs a specific \( a \)–\( i \) combination whose plane precesses 0.986° per day; most weather polar orbiters are both
Ascending vs descending nodenorthbound vs southbound equator crossing; 12 h apart in local solar time
Altitude vs orbit radius\( h \) from the surface vs \( a = R_E + h \) from Earth’s centre; Kepler’s law uses \( a \)
IFOV vs pixel spacingangular field of one detector vs distance between sample centres (they can differ)
Swath vs footprintwidth of the whole scanned strip vs size of one pixel

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper. Take \( \mu = 3.986\times 10^{14} \) m\(^3\) s\(^{-2}\) and \( R_E = 6378 \) km unless stated.

Level 1: Foundation

  1. State the four conditions an orbit must satisfy to be geostationary.
  2. Why is the inclination of a sun-synchronous orbit greater than 90°?
  3. A sun-synchronous satellite crosses the equator northbound at 13:30 local solar time. At what local time is its southbound crossing?
Answers
  1. Circular, equatorial (inclination 0°), prograde, and with a period of one sidereal day (86,164 s).
  2. The node must precess eastward at 0.986° per day. Since \( \dot\Omega \propto -\cos i \), a positive rate needs \( \cos i < 0 \), i.e. \( i > 90^\circ \) (retrograde orbit).
  3. 01:30 local solar time. The two nodes are on opposite sides of the Earth, 12 hours apart in local time.

Level 2: Part B Style

Practice Q1. The orbital period of a satellite in a circular orbit at an altitude of 850 km is closest to:

(a) 85 min (b) 102 min (c) 120 min (d) 12 h

Answer: (b) 102 min

\( a = 6378 + 850 = 7228 \) km. \( T = 2\pi\sqrt{a^3/\mu} = 2\pi\sqrt{(7.228\times 10^{6})^3/3.986\times 10^{14}} \approx 6116 \) s \( \approx 101.9 \) min. (a) is roughly the period at the surface; (d) is the GPS orbit.

Practice Q2. For the satellite in Q1, how far apart in longitude are two successive northbound equator crossings?

(a) about 14° (b) about 25° (c) about 51° (d) 0°, since the orbit is sun-synchronous

Answer: (b) about 25°

In 101.9 min the Earth turns \( 360^\circ\times 6116/86164 \approx 25.6^\circ \) (25.5° using a solar day). The ground track shifts westward by that amount, about 2,840 km at the equator. Sun-synchronism fixes the local time of crossing, not the longitude.

Practice Q3. A geostationary imager is required to give a 4 km pixel at the sub-satellite point. Its IFOV must be about:

(a) 11 μrad (b) 112 μrad (c) 1.1 mrad (d) 4.7 mrad

Answer: (b) 112 μrad

\( \text{IFOV} = d/H = 4\ \text{km}/35{,}786\ \text{km} = 1.12\times 10^{-4} \) rad. (c) is the IFOV that gives about 1 km from a polar orbiter at 850 km.

Practice Q4. Which statement about geostationary meteorological satellites is incorrect?

(a) They orbit at about 6.6 Earth radii from Earth’s centre. (b) They provide images of the same disc every 10–30 minutes. (c) They give their best coverage of the polar regions in winter. (d) Their orbital speed is about 3.1 km/s.

Answer: (c)

The polar regions lie at or beyond the limb as seen from the equatorial orbit, at satellite zenith angles too large for useful imagery, in every season. (a) 42,164/6,378 ≈ 6.6. (d) \( 2\pi\times 42{,}164/86{,}164 \approx 3.07 \) km/s.

Level 3: Part C Style

Practice Q5. A thunderstorm top at 12 km over central India is viewed by a geostationary satellite at a satellite zenith angle of 50°. A rain gauge network places the storm’s heaviest rain directly below its true position. Which statements are correct?

(A) The satellite places the cloud top about 14 km from its true position. (B) The displacement is toward the sub-satellite point. (C) The displacement would be about 10 km at a zenith angle of 45°, for a cloud top at 10 km. (D) Parallax disappears if the image is resampled to a finer grid.

(a) A and B (b) A and C (c) B and D (d) A, C and D

Answer: (b) A and C

\( \Delta x = h\tan\theta = 12\tan 50^\circ = 14.3 \) km, so (A) is correct. The line of sight from an equatorial satellite reaches the cloud top before it would reach the ground beneath it, so the cloud is mapped away from the sub-satellite point (poleward for India); (B) is wrong. \( 10\tan 45^\circ = 10 \) km, so (C) is correct. Parallax is a geometric effect of cloud height and viewing angle; regridding cannot remove it, only a correction using the cloud height can; (D) is wrong.

Practice Q6. A planner must choose satellites for three tasks: (P) tracking the hourly growth of a squall line over Odisha; (Q) measuring the temperature profile over Antarctica in July; (R) measuring how tropical oceanic rainfall varies with local time of day. Which assignment is best?

(a) P geostationary; Q sun-synchronous polar; R low-inclination precessing orbit (b) P sun-synchronous polar; Q geostationary; R sun-synchronous polar (c) P geostationary; Q geostationary; R geostationary (d) P low-inclination; Q sun-synchronous polar; R sun-synchronous polar

Answer: (a)

(P) needs frequent repeat over one region: geostationary. (Q) needs polar coverage and a sounder (often microwave, which also works in the polar night): only polar orbiters see Antarctica and carry sounders. (R) needs sampling at all local times over the tropics: a sun-synchronous satellite always passes at the same two local times, so it cannot resolve the diurnal cycle; a precessing low-inclination orbit (TRMM 35°, Megha-Tropiques 20°) can. A geostationary infrared imager could also give the diurnal cycle of cold cloud, but not with microwave or radar accuracy, which is why (c) is not best.

Practice Q7. Using \( \dot\Omega = -\tfrac{3}{2}J_2(R_E/a)^2 n\cos i \) with \( J_2 = 1.083\times 10^{-3} \), the inclination needed for a sun-synchronous orbit at 850 km altitude is closest to:

(a) 82.0° (b) 90.0° (c) 98.8° (d) 108.5°

Answer: (c) 98.8°

Required \( \dot\Omega = 2\pi/(365.24\times 86400) = 1.991\times 10^{-7} \) rad s\(^{-1}\). With \( T = 6116 \) s, \( n = 2\pi/T = 1.027\times 10^{-3} \) rad s\(^{-1}\); \( (R_E/a)^2 = (6378/7228)^2 = 0.7786 \).

\( \cos i = -1.991\times 10^{-7}/(1.5\times 1.083\times 10^{-3}\times 0.7786\times 1.027\times 10^{-3}) = -0.1532 \), so \( i = 98.8^\circ \). (a) has the right magnitude of \( \cos i \) but the wrong sign: it would precess westward.

Quick Revision Table

IdeaFact or formula
Kepler III\( T = 2\pi\sqrt{a^3/\mu} \), \( \mu = 3.986\times 10^{14} \) m\(^3\) s\(^{-2}\); \( v = \sqrt{\mu/a} \)
Geostationarycircular, equatorial, prograde, \( T = 86{,}164 \) s; \( a = 42{,}164 \) km; \( h = 35{,}786 \) km; \( v \approx 3.07 \) km/s
Geostationary viewdisc 17.4° wide; useful to about 60–70° from sub-point; eclipses near equinoxes up to about 70 min
Polar orbiter700–900 km; about 100–102 min; about 14 orbits a day; tracks about 25° apart
Sun-synchronous\( \dot\Omega = 0.986^\circ \) per day eastward; \( i \approx 98^\circ \)–\( 99^\circ \); fixed local equator-crossing time
Low inclinationTRMM 35°, GPM 65°, Megha-Tropiques 20°; sample the diurnal cycle
Footprint\( d = H\times\text{IFOV} \); grows off nadir
Parallax\( \Delta x = h\tan\theta \), away from the sub-satellite point
Indian seriesINSAT, Kalpana-1 (2002), INSAT-3D (2013) with imager and sounder, 3DR, 3DS; data via MOSDAC
First weather satelliteTIROS-1, 1 April 1960

Key Takeaways

What’s Next