Lecture 3d: Atmospheric adia4on & emote Sensing n
Outline 1. adar Basics & adar Equation 2. adar eflectivity & Precipitation 3. adar Attenuation 4. Lidar Basics & Lidar Equation 5. Lidar emote Sensing
The radar equa4on: P r = C K 2 2 n(d) D 6 dd = C K 2 2 Z It assumes that empty space exists between the radar and the illuminated backscabering volume. But in reality, abenua4on occurs due to, for example, absorp4on by atmospheric gases, absorp4on by cloud droplets, or scabering and absorp4on by precipita4on. The abenua4on can be expressed using the Beer s Law: ΔP t = 2σ ext P t Δr (This is basically Beer s Law. Why is there a factor of 2?) P r = C K 2 Z exp( 2 σ 2 ext dr) dbz e =1log 1 (Ze 2 σ ext dr ) =1log 1 Z 1log 1 e (Z e : Effec4ve reflec4vity) 2σ ext dr
dbz e =1log 1 Z 1log 1 e 2σ ext dr ABenua4on per unit distance σ decadic ext 1(log 1 e)σ ext = k c w w: density of the material that cause ex4nc4on (e.g., water droplets inside clouds) ABenua4on is: 1) a weak func4on of temperature, 2) strong func4on of wavelength, 3) strong func4on of phase of H 2 O x 1-3
CloudSat: 3-mm (or 94 GHz) dbz e =1log 1 Z 1log 1 e 2σ ext dr Example of heavy abenua4on: the original Z should be very large because of heavy precipita4on
3mm wavelength radar 2cm wavelength radar
Outline 1. adar Basics & adar Equa4on 2. adar eflec4vity & Precipita4on 3. adar ABenua4on 4. Lidar Basics & Lidar Equa4on 5. Lidar emote Sensing
Fundamentally, lidar is nothing different from radar except using different EM spectrum (visible/i). adar equa4on: P r = C K 2 Z exp( 2 σ 2 ext dr) (Key assump4on: ayleigh scabering) where Z n(d) D 6 dd Lidar Lidar equa4on: P r = C 2 h 2 β 4π exp( 2 σ ext dr) β is backscabering coefficient: km -1 ster -1. h is the length of the emibed pulse. C: Lidar constant (hiding all other parameters) β (backscaber coefficient) is the cousin of the radar reflec4vity, Z. However, it s not readily derived from scabering theory since backscabering at lidar wavelengths oden arises from scabering by a variety of different size par4cles such as air molecules, aerosols, and cloud par1cles (i.e., mixture of ayleigh & Mie).
Lidar Equa4on P r = C 2 h 2 β 4π exp( 2 σ ext dr) In Lidar community, they define a new parameter called signal variable (S): S() ln( 2 P r ) S S = ln( β β ) 2 σ ext dr where β is the reference value at One notable different between radar & lidar is abenua4on: lidar is abenuated much faster (note the exponen4al decrease of the signal).
The A-Train CloudSat: a 94-GHz cloud profiling radar CALIPSO: dual-channel lidar + I radiometer A-Train stands for Adernoon Train, referring to the equator crossing 4me of the sun-synchronous orbit (1:3 pm ascending)
Lidar is used for thin clouds and aerosols adar is used for detec4ng thick cloud & precipita4on
Lidar Equa4on: P r = C 2 h 2 β 4π exp( 2 σ ext dr) β is backscabering coefficient: km -1 ster -1. h is the length of the emibed pulse. C: Lidar constant (hiding all other parameters) Define signal variable (S) S S = ln( β β ) 2 σ ext dr Deriva4ve of S w.r.t. range () ds d = 1 dβ β d 2σ ext S is what we measure, and it depends on two variables: backscabering coefficient (β) and ex4nc4on coefficient (σ ext ). The simplest assump4on: dβ/d=, i.e., scaberers are homogeneously distributed with range along the lidar path. More generally, we can assume the backscabering ability and ex4nc4on ability are related: β = bσ ext n, where b, n are specified const. ds d = n σ ext dσ ext d 2σ ext
High Spectral esolu9on Lidar (HSL) Up to this point, we have not touched upon the issue of the spectral width of the pulse. In reality, the received pulse is broadened due to the random mo4ons of the scaberers. The broadening of the heavier, more slowly moving par4cles (e.g., aerosols) is much narrower than that of the lighter and fast-moving ones (e.g., molecules). HSL uses a combina4on of a tunable laser, and a very narrow spectral filter to disentangle the contribu4on to the backscabering by molecules from aerosols.
Differen9al Absorp9on Lidar (DIAL) DIAL exploits the strong wavelength dependence of gaseous absorp4on. P r = C 2 h 2 β 4π exp( 2 σ ext dr) DIAL chooses two wavelengths, λ 1 and λ 2, that are close enough such that the differences in par4cle op4cal proper4es, namely, σ ext and β, are small. But difference in k is large if λ 1 and λ 2 are along the slope of the gaseous absorp4on spectrum. β is backscabering coefficient: km -1 ster -1. h is the length of the emibed pulse. C: Lidar constant (hiding all other parameters) ln P r,1 P r,2 2 N(k 1 k 2 )dr The ex4nc4on can be expressed as: σ ext = σ ext,aerosol + Nk gas N is number concentra4on of the absorbing molecule and k is the absorp4on coefficient k 1 and k 2 are known from molecular spectroscopy, so N - gas concentra4on - can be retrieved from a two-wavelength lidar system.