Skip to content

Weather Radar and Rainfall Measurement

CSIR NET Part C meteorology: radar equation, reflectivity and dBZ, Z–R relations, bright band, Doppler and dual-pol radar, rain gauges and areal rainfall.

  • 21 min read
  • Updated
On this page · 19 sections
  1. Prerequisites
  2. The Hook: The Clutter That Became a Science
  3. How a Weather Radar Works
  4. The Radar Equation and Reflectivity
  5. From Reflectivity to Rain Rate
  6. The Bright Band
  7. Beam Propagation
  8. Doppler Radar
  9. Dual-Polarisation Radar
  10. Radar Displays and Signatures
  11. Measuring Rainfall at the Ground
  12. Radar–Gauge Merging and Satellite Estimates
  13. Memory Tricks
  14. Don’t Confuse
  15. Common Mistakes
  16. Practice
  17. Quick Revision Table
  18. Key Takeaways
  19. What’s Next

Prerequisites

The Hook: The Clutter That Became a Science

How a Weather Radar Works

A radar transmits short pulses of microwave energy through a directional antenna and listens for the echo between pulses.

QuantityRelationMeaning
Range\( r = ct/2 \)\( t \) is the round-trip time
Range resolution\( c\tau/2 \)\( \tau \) is the pulse duration; 1 μs gives 150 m
Pulse repetition frequency (PRF)pulses per secondsets maximum range and maximum velocity
Beamwidthabout \( \lambda/D_{ant} \) (radians)a 1° beam is about 1.7 km wide at 100 km
Sample volumebeam cross-section × \( c\tau/2 \)grows as \( r^2 \)

Frequency bands

BandFrequencyWavelengthUse and limitation
S2–4 GHzabout 10 cmlong-range precipitation and severe weather; little attenuation; large antenna needed
C4–8 GHzabout 5 cmcommon national networks; moderate attenuation in heavy rain
X8–12 GHzabout 3 cmshort-range, mobile and research radars; strong attenuation
Ku, Ka12–40 GHzabout 2 cm to under 1 cmspaceborne precipitation radars; airborne
Wabout 94 GHzabout 3 mmcloud radars: sense cloud droplets; heavily attenuated by rain

Shorter wavelengths see smaller particles (backscatter \( \propto \lambda^{-4} \)) but are attenuated more. Longer wavelengths penetrate heavy rain but need large antennas for the same beamwidth.

The Radar Equation and Reflectivity

Point target versus distributed target

For a single target of backscattering cross-section \( \sigma \),

$$P_r = \frac{P_tG^2\lambda^2\sigma}{(4\pi)^3r^4}$$

Precipitation fills the beam. The number of scatterers in the sample volume grows as \( r^2 \), cancelling two powers of \( r \). The result for beam-filling precipitation is

$$\boxed{P_r = \frac{C\,|K|^2\,Z}{r^2}}$$

where \( C \) collects the radar constants (transmitted power, antenna gain, beamwidths, pulse length, wavelength), \( |K|^2 \) depends on the dielectric constant of the scatterers and \( Z \) is the radar reflectivity factor. Precipitation echo power falls as \( 1/r^2 \), not \( 1/r^4 \).

Rayleigh scattering and the reflectivity factor

When drops are small compared with the wavelength (\( D \ll \lambda \), true for rain at S and C band), the backscattering cross-section of a sphere is

$$\sigma_b = \frac{\pi^5|K|^2D^6}{\lambda^4}$$

Summing over the drops in a unit volume gives

$$\boxed{Z = \sum_{\text{unit volume}} D^6 = \int_0^\infty N(D)\,D^6\,dD}$$

conventionally in mm\(^6\) m\(^{-3}\), and expressed logarithmically:

$$\boxed{\text{dBZ} = 10\log_{10}\left(\frac{Z}{1\ \text{mm}^6\ \text{m}^{-3}}\right)}$$
ValueMeans
(K
(K
Equivalent reflectivity \( Z_e \)the \( Z \) that liquid drops would need to give the observed power; what radars actually report

Because ice has a \( |K|^2 \) about 4.7 times smaller than water (about 7 dB), dry snow gives a weaker echo than rain of the same mass.

From Reflectivity to Rain Rate

Drop-size distributions

The Marshall–Palmer distribution, fitted to measured raindrop spectra, is exponential:

$$\boxed{N(D) = N_0\,e^{-\Lambda D}, \qquad N_0 = 8000\ \text{m}^{-3}\,\text{mm}^{-1}, \qquad \Lambda = 4.1\,R^{-0.21}\ \text{mm}^{-1}}$$

with \( R \) in mm h\(^{-1}\). Heavier rain has a smaller \( \Lambda \): relatively more large drops. Modern work often uses a gamma distribution \( N(D) = N_0D^\mu e^{-\Lambda D} \), which can represent the deficit of very small drops.

The integrals that connect the distribution to measurable quantities (with \( D \) in mm, \( v \) in m s\(^{-1}\), \( N \) in m\(^{-3}\) mm\(^{-1}\)):

$$Z = \int N D^6\,dD, \qquad R = 6\pi\times 10^{-4}\int N D^3 v(D)\,dD\ \ (\text{mm h}^{-1}), \qquad W = \frac{\pi}{6}\rho_w\int N D^3\,dD$$

\( Z \) is the sixth moment, liquid water content the third, and rain rate roughly the 3.67th (since \( v \propto D^{0.67} \) approximately). Because they are different moments, no single \( Z \)–\( R \) relation is exact.

Z–R relations

$$\boxed{Z = aR^b}$$
RelationTypical use
\( Z = 200R^{1.6} \)Marshall–Palmer; stratiform rain
\( Z = 300R^{1.4} \)default convective relation of the US NEXRAD network
\( Z = 250R^{1.2} \)tropical rain with many small drops

With \( Z = 200R^{1.6} \):

dBZ\( R \) (mm h\(^{-1}\))Description
200.65light
302.7moderate
4011.5heavy
4523.7very heavy
5048.6intense
55100extreme; hail likely contaminates above this

Inverting: \( R = (Z/200)^{1/1.6} \). Two useful consequences: doubling \( R \) adds \( 16\log_{10}2 \approx 4.8 \) dB; a tenfold increase in \( R \) adds 16 dB.

Errors in radar rainfall

Error sourceEffect
Drop-size variabilitythe single biggest uncertainty: same \( Z \), different \( R \)
Calibrationa 1 dB error in \( Z \) gives about a 15% error in \( R \) (with \( b = 1.6 \))
Bright bandstrong overestimate where the beam passes through the melting layer
Hailhuge \( Z \) from few large stones; rain rate grossly overestimated (rates are often capped)
Beam overshooting and broadeningat long range the beam passes above shallow rain or samples only its top
Beam blockagemountains or buildings block part of the beam
Attenuationat C and X band, heavy rain weakens the beam behind it
Anomalous propagationsuper-refraction bends the beam into the ground: false echoes
Evaporation or growth below the beamthe rain measured aloft is not what reaches the ground

The Bright Band

In stratiform precipitation, a horizontal layer of enhanced reflectivity appears a few hundred metres below the 0 °C level.

graph TD
    A[Above 0 °C level: dry snowflakes, low K squared, moderate Z] --> B[Melting layer: flakes acquire a water coat]
    B --> C[Water-coated flakes: large size plus liquid K squared: Z peaks]
    C --> D[Below: flakes collapse to drops, smaller and falling faster]
    D --> E[Fewer particles per volume and smaller sizes: Z decreases]
  1. Above: dry snow has \( |K|^2 \approx 0.2 \).
  2. In the melting layer: a film of water gives the large, still-unmelted flakes the dielectric factor of water, so \( Z \) jumps by roughly the ratio of the dielectric factors (about 7 dB) and more, because the flakes are big and may aggregate.
  3. Below: the flakes collapse into compact raindrops that are smaller and fall several times faster (about 1 m s\(^{-1}\) for snow against several m s\(^{-1}\) for rain). Faster fall means fewer particles per unit volume, and \( Z \) falls again.

Consequences: the bright band marks the melting level (useful for forecasting the snow level); it identifies stratiform rain and the ice process (no ice, no melting layer); and it is a major source of overestimation in radar rainfall. Convective rain, with strong updrafts mixing particles vertically, usually shows no clear bright band.

Beam Propagation

The beam is refracted downward by the normal decrease of refractive index with height. The standard treatment replaces the Earth’s radius \( a \) by an effective radius \( k_ea \) with \( k_e = 4/3 \), and draws the beam as a straight line:

$$\boxed{h = \sqrt{r^2 + (k_ea)^2 + 2rk_ea\sin\theta_e} - k_ea \approx r\sin\theta_e + \frac{r^2}{2k_ea}}$$

with \( k_ea \approx 8500 \) km. At an elevation of 0.5°, the beam centre is about 1.5 km above the radar at 100 km range and about 4.1 km at 200 km. At long range, low-level rain and shallow clouds are simply not seen.

When the refractive index decreases faster than normal (a strong inversion with dry air above moist air), the beam bends more (super-refraction) and may be trapped (ducting), giving ground clutter at long range: anomalous propagation. See Atmospheric Optics for the refractivity conditions.

Doppler Radar

A Doppler radar measures the phase change of the echo from pulse to pulse, which gives the radial velocity \( v_r \) of the scatterers (toward or away from the radar):

$$\boxed{f_D = \frac{2v_r}{\lambda}}$$

A 20 m s\(^{-1}\) radial velocity at 10 cm wavelength gives \( f_D = 400 \) Hz.

Range and velocity ambiguity

LimitFormulaReason
Maximum unambiguous range\( r_{max} = \dfrac{c}{2\,\text{PRF}} \)an echo arriving after the next pulse is sent is assigned the wrong range (“second-trip echo”)
Maximum unambiguous (Nyquist) velocity\( v_N = \pm\dfrac{\lambda\,\text{PRF}}{4} \)the phase change between pulses is ambiguous beyond \( \pm\pi \)
Doppler dilemma\( r_{max}\,v_N = \dfrac{c\lambda}{8} \)a fixed product: increasing one decreases the other
\( \lambda \), PRF\( r_{max} \)\( v_N \)
10 cm, 1000 Hz150 km±25 m s\(^{-1}\)
5 cm, 1000 Hz150 km±12.5 m s\(^{-1}\)
10 cm, 500 Hz300 km±12.5 m s\(^{-1}\)

Longer wavelengths ease the dilemma, which is why S-band radars are preferred for severe weather and cyclones. A velocity beyond \( v_N \) folds (aliases): with \( v_N = 15 \) m s\(^{-1}\), a true 20 m s\(^{-1}\) appears as \( 20 - 2(15) = -10 \) m s\(^{-1}\). Dual-PRF techniques and continuity-based software unfold such values.

What Doppler fields show

  • Mesocyclone / rotation: a tight couplet of inbound and outbound velocities side by side.
  • Convergence / divergence: inbound and outbound along the same radial (divergence near the ground under a downburst).
  • Velocity–azimuth display (VAD): on a conical scan in uniform wind, radial velocity varies sinusoidally with azimuth; the amplitude and phase give the wind speed and direction at each height, a vertical wind profile from one radar.
  • Spectrum width: the spread of velocities in the sample volume, a measure of turbulence and shear.
  • Doppler radars also filter ground clutter, whose velocity is zero.

Dual-Polarisation Radar

Transmitting and receiving horizontally and vertically polarised waves reveals particle shape, phase and diversity:

VariableDefinitionWhat it shows
Differential reflectivity \( Z_{DR} \)\( 10\log_{10}(Z_H/Z_V) \)positive for oblate raindrops, increasing with drop size; near 0 dB for tumbling hail and small drops
Specific differential phase \( K_{DP} \)rate of change of differential phase with range (° km\(^{-1}\))liquid water content in heavy rain; unaffected by calibration and by attenuation
Co-polar correlation coefficient \( \rho_{hv} \)correlation of H and V returnsclose to 1 in pure rain or snow; lower in the melting layer (mixed phase); much lower for non-meteorological echoes (birds, insects, debris)

Uses: hydrometeor classification (rain, snow, graupel, hail, melting layer), better rain-rate estimates in heavy rain from \( K_{DP} \), hail detection (high \( Z \) with low \( Z_{DR} \)) and removal of non-weather echoes.

Radar Displays and Signatures

DisplayWhat it shows
PPI (plan position indicator)one elevation angle, all azimuths: a map
RHI (range–height indicator)one azimuth, many elevations: a vertical section
CAPPI (constant-altitude PPI)a horizontal slice at fixed height interpolated from several elevations
Echo top, VILmaximum height of a reflectivity threshold; vertically integrated liquid

Signatures worth recognising: the hook echo at the rear flank of a supercell (rotation), the bounded weak echo region under a strong updraft, bow echoes with damaging winds, and the bright band in stratiform rain.

Measuring Rainfall at the Ground

Rain gauges

Rainfall is reported as a depth: the depth of water that would stand on a flat surface. \( 1 \) mm of rain \( = 1 \) L m\(^{-2}\).

$$\boxed{\text{depth} = \frac{\text{collected volume}}{\text{collector area}}}$$

500 mL collected in a gauge with a 200 cm\(^2\) rim is \( 500/200 = 2.5 \) cm \( = 25 \) mm.

GaugePrincipleNotes
Ordinary (non-recording) gaugefunnel and collector, read once or twice a day with a graduated measuring jarthe network standard for daily totals
Tipping-bucket gaugea small two-compartment bucket tips after a fixed depth (commonly 0.2 or 0.5 mm), each tip recordedgives intensity; undercounts very heavy rain because water is lost while the bucket tips
Weighing gaugeweighs the accumulated precipitation continuouslyhandles snow and hail; used in automatic stations
Float (siphon) recordera float rises with collected water and drives a pen; siphons empty the chamber at a fixed leveltraditional self-recording gauge giving a mass curve
Disdrometeroptical or impact sensor counting and sizing individual dropsmeasures the drop-size distribution itself, linking gauges and radar

Gauge errors

  • Wind-induced undercatch: wind speeds up over the gauge rim and deflects drops; the largest error, worse for snow and small drops. Wind shields and pit gauges reduce it.
  • Wetting and evaporation losses from the funnel and collector.
  • Splash in or out, and siting: gauges must be well clear of trees and buildings, at a standard rim height.
  • Representativeness: a gauge samples a few hundred square centimetres; convective rain varies over hundreds of metres.

Areal rainfall from a gauge network

MethodIdeaWhen
Arithmetic meansimple average of gaugesflat terrain, uniform and dense network
Thiessen polygonsweight each gauge by the area of its polygon (perpendicular bisectors between neighbouring gauges)uneven gauge spacing; ignores orography
Isohyetal methoddraw lines of equal rainfall; area-weight the mean between successive isohyetsthe most accurate when drawn with knowledge of terrain and storm structure
$$\bar P_{\text{Thiessen}} = \frac{\sum_i P_iA_i}{\sum_i A_i}$$

Radar–Gauge Merging and Satellite Estimates

Radar–gauge merging: radar gives the spatial pattern, gauges the absolute value. The simplest adjustment is the mean field bias \( B = \sum G_i/\sum R_i \) over gauge locations, applied to the whole radar field; more advanced methods let the bias vary in space.

Satellite rainfall estimation:

ApproachPrincipleLimitation
Infrared (IR)cold, high cloud tops associated with deep convection; for example the GOES Precipitation Index assigns 3 mm h\(^{-1}\) to pixels colder than 235 Kindirect; misses warm rain; cirrus gives false rain
Passive microwaveemission from liquid water (best over the radiometrically cold ocean); scattering by ice at high frequencies (usable over land)coarse resolution; sparse overpasses from low orbit
Spaceborne radarmeasures reflectivity profiles directly: TRMM Precipitation Radar (Ku band, 1997–2015); GPM Dual-frequency Precipitation Radar (Ku and Ka bands, launched 2014)narrow swath; infrequent sampling
Merged productscombine microwave, IR and gauges (for example GPM-era multi-satellite products)blend of all the above errors

India’s geostationary INSAT-series imagers provide IR-based rainfall estimates over the region, and the Indo-French Megha-Tropiques mission was dedicated to the tropical water cycle.

Memory Tricks

Don’t Confuse

ThisvsThat
\( P_r \propto 1/r^4 \): point target\( P_r \propto 1/r^2 \): beam-filling precipitation
\( Z \): sixth moment of the DSDrain rate: about 3.67th moment; LWC: third moment
\( Z_e \): what the radar reports (water (K^2 ) assumed)
\( r_{max} = c/2\text{PRF} \)\( v_N = \lambda\,\text{PRF}/4 \)
\( Z_{DR} \) near 0 for hail (tumbling)\( Z_{DR} \) positive for big raindrops (oblate)
Bright band: stratiform rain, melting layerconvective rain: usually no clear bright band
Tipping bucket: intensity, undercounts heavy rainweighing gauge: handles snow, total mass
IR rain estimates: cloud-top temperaturemicrowave: liquid emission and ice scattering

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. Define the radar reflectivity factor \( Z \).
  2. How does the echo power from beam-filling precipitation depend on range?
  3. What is the Doppler shift for a radial velocity \( v_r \) at wavelength \( \lambda \)?
Answers
  1. \( Z = \sum D^6 \) per unit volume (mm\(^6\) m\(^{-3}\)), valid for Rayleigh scatterers.
  2. As \( 1/r^2 \).
  3. \( f_D = 2v_r/\lambda \).

Level 2: Part B Style

Practice Q1. Using \( Z = 200R^{1.6} \), a reflectivity of 45 dBZ corresponds to a rain rate closest to:

(a) 5 mm h\(^{-1}\) (b) 12 mm h\(^{-1}\) (c) 24 mm h\(^{-1}\) (d) 50 mm h\(^{-1}\)

Answer: (c) 24 mm/h

\( Z = 10^{4.5} = 31{,}623 \); \( R = (31{,}623/200)^{1/1.6} = (158.1)^{0.625} \approx 23.7 \) mm h\(^{-1}\).

Practice Q2. A C-band Doppler radar has \( \lambda = 5 \) cm and PRF 1200 Hz. Its maximum unambiguous range and Nyquist velocity are:

(a) 125 km, ±15 m s\(^{-1}\) (b) 250 km, ±15 m s\(^{-1}\) (c) 125 km, ±30 m s\(^{-1}\) (d) 250 km, ±60 m s\(^{-1}\)

Answer: (a)

\( r_{max} = 3\times 10^8/(2\times 1200) = 125 \) km; \( v_N = 0.05\times 1200/4 = 15 \) m s\(^{-1}\).

Practice Q3. Using the 4/3 effective Earth radius (\( k_ea \approx 8500 \) km), the height of the beam centre of a 0.5° elevation scan at 200 km range is closest to:

(a) 1.7 km (b) 2.4 km (c) 4.1 km (d) 6.0 km

Answer: (c) 4.1 km

\( h \approx r\sin\theta + r^2/(2k_ea) = 200\times 0.00873 + 200^2/17{,}000 = 1.75 + 2.35 = 4.1 \) km. Option (a) ignores curvature; option (b) keeps only curvature.

Practice Q4. A tipping-bucket gauge of 0.2 mm per tip records 60 tips in 30 minutes. The mean rain intensity over that period is:

(a) 6 mm h\(^{-1}\) (b) 12 mm h\(^{-1}\) (c) 24 mm h\(^{-1}\) (d) 36 mm h\(^{-1}\)

Answer: (c) 24 mm/h

\( 60\times 0.2 = 12 \) mm in 0.5 h, so 24 mm h\(^{-1}\). Option (b) forgets to convert to an hourly rate.

Level 3: Part C Style

Practice Q5. Four gauges in a catchment of 400 km\(^2\) record storm totals, and their Thiessen polygon areas are:

GaugeRainfall (mm)Polygon area (km\(^2\))
A42120
B5880
C35150
D7050

Which are correct?

  1. The arithmetic mean rainfall is 51.25 mm.
  2. The Thiessen mean rainfall is about 46.1 mm.
  3. The Thiessen mean is lower because the gauges with the most rain represent the smallest areas.
  4. The volume of rain over the catchment by the Thiessen method is about \( 1.8\times 10^{7} \) m\(^3\).

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 2 and 4 only

Answer: (c) 1, 2, 3 and 4
  • 1 true: \( (42 + 58 + 35 + 70)/4 = 51.25 \) mm.
  • 2 true: \( (42\times 120 + 58\times 80 + 35\times 150 + 70\times 50)/400 = (5040 + 4640 + 5250 + 3500)/400 = 18{,}430/400 = 46.1 \) mm.
  • 3 true: D (70 mm) and B (58 mm) have the smallest polygons; C (35 mm) the largest.
  • 4 true: \( 0.0461 \) m \( \times 4\times 10^8 \) m\(^2\) \( = 1.84\times 10^7 \) m\(^3\).

Practice Q6. At three gauges, storm totals are 12, 20 and 8 mm, while the radar (using \( Z = 200R^{1.6} \)) estimates 9, 16 and 5 mm at the same points. Assume the whole discrepancy comes from a radar calibration error. Which are correct?

  1. The mean field bias \( \sum G/\sum R \) is about 1.33.
  2. The radar underestimates rainfall by about 1.25 dB in rain-rate terms.
  3. The reflectivity calibration is about 2 dB too low.
  4. Multiplying every radar pixel by the mean field bias removes errors caused by drop-size variability within the storm.

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2 and 4 only (d) 1, 2, 3 and 4

Answer: (b) 1, 2 and 3 only
  • 1 true: \( 40/30 = 1.333 \).
  • 2 true: \( 10\log_{10}1.333 = 1.25 \) dB.
  • 3 true: \( R \propto Z^{1/1.6} \), so a factor 1.333 in \( R \) needs a factor \( 1.333^{1.6} = 1.585 \) in \( Z \), i.e. \( 10\log_{10}1.585 \approx 2.0 \) dB.
  • 4 false: a single multiplicative factor corrects a uniform bias (such as calibration) but not spatially varying errors such as different drop-size distributions in different parts of the storm.

Practice Q7. A vertically pointing radar in stratiform rain records this reflectivity profile; the 0 °C level from a radiosonde is at 4.6 km:

Height (km)6.05.04.44.23.83.02.0
dBZ22253336292828

Which statements are correct?

  1. The peak at 4.2–4.4 km is a bright band caused by melting snow.
  2. Using the 4.2 km value in a Z–R relation would overestimate the surface rain rate.
  3. The fall speed of the hydrometeors increases from above to below the peak.
  4. The presence of this feature shows that precipitation formed entirely by collision–coalescence.

(a) 1, 2 and 3 only (b) 1 and 4 only (c) 2 and 3 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 3 only
  • 1 true: a reflectivity peak just below the 0 °C level, with lower and nearly constant values beneath, is the bright band.
  • 2 true: 36 dBZ gives about 6.5 mm h\(^{-1}\) with \( Z = 200R^{1.6} \), against about 2.1 mm h\(^{-1}\) from the 28 dBZ in the rain below.
  • 3 true: snowflakes fall at about 1 m s\(^{-1}\); raindrops at several m s\(^{-1}\). That speed-up is one reason \( Z \) falls below the peak.
  • 4 false: a melting layer requires ice above it, so the ice process contributed. (The profile is from practice data, not an observed case.)

Quick Revision Table

IdeaFact or formula
Range, resolution\( r = ct/2 \); \( c\tau/2 \)
BandsS about 10 cm, C 5 cm, X 3 cm; Ku and Ka spaceborne; W 94 GHz cloud radar
Radar equationpoint target \( 1/r^4 \); precipitation ( P_r = C
Reflectivity\( Z = \sum D^6 \) (mm\(^6\) m\(^{-3}\)); dBZ \( = 10\log Z \); (
Marshall–Palmer\( N_0 = 8000 \) m\(^{-3}\) mm\(^{-1}\); \( \Lambda = 4.1R^{-0.21} \) mm\(^{-1}\); \( Z = 200R^{1.6} \)
Z–R markers30 dBZ ≈ 2.7, 40 ≈ 11.5, 50 ≈ 49 mm h\(^{-1}\); ×2 in \( R \) = +4.8 dB
Bright bandbelow 0 °C level; water-coated snow; stratiform; overestimates rain
Beam height\( h \approx r\sin\theta + r^2/2k_ea \); \( k_e = 4/3 \)
Doppler\( f_D = 2v_r/\lambda \); \( r_{max} = c/2\text{PRF} \); \( v_N = \lambda\text{PRF}/4 \); product \( c\lambda/8 \)
Dual-pol\( Z_{DR} \) shape; \( K_{DP} \) liquid water; \( \rho_{hv} \) mixture or non-weather
Gaugesordinary, tipping bucket (0.2 or 0.5 mm per tip), weighing, siphon, disdrometer
Gauge errorswind undercatch (largest), wetting, evaporation, splash
Areal rainfallarithmetic, Thiessen, isohyetal
SatellitesIR (GPI: 3 mm h\(^{-1}\) below 235 K), passive microwave, TRMM PR, GPM DPR

Key Takeaways

What’s Next

Advertisement
Advertisement