Freezing contamination : aircraft icing

Freezing contamination : aircraft icing

Total duration (4 hours)

Ross Paulson photography

Freezing contamination :aircraft icing

Aviation meteorology : significant phenomena - freezing contamination- aircraft icing

2

(Description of the hazard)

A few definitions (5) Process of ice formation (6)An historic accident (7)Statistics on aircraft accidents related to icing (8)Costs induced by icing (9)

General framework for this presentation (10) (Effects on aircrafts)First part : aviation and icing (11)

Different types of accretion (12-15)Intensity of ice accretion (16-17)Consequences of accretion (18-19)Vulnerability factors examples (20-21)Specific vulnerabilities (22-24)Detection in flight (25)Removal of accretion : De-icing (26)Prevention of icing : Anti-icing (27)Certification and icing conditions (28)Marginal weather conditions (29)

Conclusion of the first part : a list with meteorological objectives on icing (30)

Contents (1)


3

(Diagnosis, forecasting)Second part : meteorology and icing (31)

Prognostic variables in the atmosphereAir temperature (including 0°C level definition) (32-35)Relative humidity (36-37)Liquid Water Content (38-43)Size of the elements (44-45)Vertical velocity (46-47)Vertical instability (48)Wind shear (48)

Conclusions about icing potential (49)Favourable scenarios for icing

Convective (50-52)Turbulent (53-55) Frontal (56-57)Orographic (58-60)Freezing fog (61)

Remote sensing observation systemssatellite observation (1.6-3.9-8.7-10.8 microns and composite imagery) (62-67)radar observation (68-69)

Contents (2)


4

Forecasting techniques (70)Extrapolation of observational icing data (71-72)Looking for icing scenarios (73)Using numerical model outputs (74-75)Crossing observations with model outputs (data fusion, icing index...) (76-77)Towards a conclusion (78)

Putting in place icing conditions forecasts (79-80)

A transmission chain of adapted information (81-84)SIGMET, AIRMET, GAMET, SIGWX chartsPIREPS

Awareness of the icing phenomena (85-86)

A few points on where to go with research and development (87)

A few bibliography, internet ressources (88-89) List of key words (90)

Case study (91-)

(*) to read in the comment page

Contents (3)


5

Contamination, contaminant : general term used in aeronautics to identify the deposit of a “foreign” body onto a surface. For instance, water on a runway is a contaminant.

Icing : contamination of an aircraft by ice. This contamination can happen both on surface and in the air.

Accretion (of ice) : accumulation (of ice) on a certain part of the aircraft.

Icing Conditions : state of the atmosphere favourable for ice accretion on an aircraft.

Icing Potential : information about the state of the atmosphere giving (and graduating) the risk for icing.

A few definitions


6

Ice can form on objects in the atmosphere :

• Through freezing of liquid water drops that were deposited in the form of droplets, drops, film or puddles on an object. The temperature of the object determines the solidification.

•Through sublimation, water vapour transforms directly into ice. The amount of water is relatively small.

In an atmosphere with a sub-zero temperature :

•Through ending the state of supercooling of drops and droplets present in the atmosphere. Supercoooling is frequently present in clouds with negative temperature until –20°C.

Process of ice formation


7

On 31 October 1994, a commuter departs from Indianapolis going to Chicago-O’Hare (USA) for a regular flight taking about 1 hour 5 minutes. At 50 Nm from Chicago, the airplane is put in a holding pattern at 10 000ft because of the heavy traffic. 30 minutes later the plane crashes in a field close to Roselawn, Indiana. The 68 passengers and crew die.

Conclusions of the investigations :•Holding pattern with liquid precipitation during 24 minutes. •Air temperature varying between –2°C and –4°C. •Total estimated ice accretion between 35 and 65 millimetres. •Flight above certification limits.•No action was taken by the crew in order to “de-ice” the aircraft.•Icing conditions were communicated in the Met Flight Folder handed to the crew, through the transmission of an AIRMET signalling precipitation with temperatures below freezing at the chosen flight level of the aircraft.

An historic accident


8

•The Investment Analysis and Operations Research department of the FAA (Federal Aviation Administration) researched a little over 10 000 accidents between 1982 et 2001.

•In 20% of the cases, meteorological factors were the main cause and within this about 25% of the accidents resulted in loss of human life.

•Icing including flight in icing conditions, freezing fog and carburettor icing for light aviation represented about 10% of the meteorological factors causing these accidents.

•13% of the accidents due to icing resulted in loss of life, about 30 cases in 20 years.

Statistics on aircraft accidents related to icing


9

•On the aircraft: when designing aircraft with possibility to fly in icing conditions, it is necessary to include equipment preventing icing or de-icing equipment.

•At the airports: specific means to de-ice surfaces have to be deployed.

The methods have a high “operational” cost. •In flight the possible pay-load and autonomy are reduced. (*).

•On surface, the operational procedures are often time consuming and induce

frequent delays on the bigger airports. (**).

Costs caused by icing


10

Atmospheric Icing Potential.The meteorologist puts forward a diagnostic on the icing potential of the atmosphere. He

qualifies, localises and quantifies the phenomena using different parameters (temperature, LWC…).

The information has to be adapted for use by the aviation community

The feedback of the aviators is necessary toimprove the forecasts.

The user filters the information and the operator evaluates the risk. The “filter” consists of aircraft, pilot, ATC, airport officials or more in general

terms “the aeronautical system”. The induced effects

lead to operational decisions, otherwise the safety is compromised.

General framework for this presentation


11

Source NASA-Lewis Research Center

First part : Aviation and icing


12

Aspect : crystalline in the form of scales, needles or feathers.Formation conditions : Sublimation of water vapour into ice. This deposit can occur without clouds.Effects : Even if the amount of deposited material is low, it can be significant

under certain conditions.

hoar frost

source Transport Canada

Different types of accretion (1)


13

Aspect : opaque and white, but rather fragile brittle. Formation conditions : On a cold surface in a homogeneous cloud environment (t<<0°C). The supercooled cloud droplets rapidly freeze resulting in entrapped bubbles of air.Effects : Forms on leading edges. Rime ice is always significant and must be removed.

rime ice

Source Météo France



14

Aspect : transparent, homogeneous et smooth, very compact. Its specific mass is close to the one of pure ice.Formation conditions : On a cold surface in a homogeneous cloud environment with a temperature close to 0°C. The supercooled cloud droplets are present in large quantities and spread out before they slowly freeze.Effects : Develops in cones on the leading edges and is very significant. Should be prevented from forming.

clear ice

Source NASA-Lewis Research Centre



15

Aspect : Mix of clear ice, hoar frost and rime ice. Whitish and brittle.Formation conditions : On a cold surface in a heterogeneous cloud environment where the temperature and cloud drops sizes fluctuate (*). Effects : Similar to rime ice.

mixed ice

Source NASA-Lewis Research Centre



16

light : > 1g/cm²/hour

moderate : > 6g/cm²/hour

severe :> 12g/cm²/hour

Intensity of ice accretion (1)


17

Light Icing : does not pose any specific restraints on the behaviour of the aircraft

Moderate icing : icing conditions may cause the crew to change heading or altitude

Severe icing : icing conditions which force the crew to immediately change heading or altitude

Intensity of ice accretion (2)


18

• The accumulation of ice represents an increase in mass and leads to the modification of the longitudinal equilibrium of the aircraft. The effect is relatively small on larger aircraft.

• Icing on tubes and antennas disturbs their operation and can lead to the rupture of elements.

• Icing on the windshield reduces the visibility

• Means of propulsion (motors, propellers, fans, rotors) are also vulnerable to ice accretion. Their efficiency is reduced and they can stop functioning altogether.

Consequences of accretion (1)


19

source NASA

Lift force (*) reduces considerably (20% - 30%) when modern wings get contaminated

Moreover, forms of light icing have a similar effect as forms of severe icing.

clean

with contamination

• The aerodynamic consequences: a major impact.

Consequences of accretion (2)


20

PROFILE

In the same icing conditions, the resulting accretion and the effect on

aerodynamics vary largely from one type of aircraft to another.

The aerodynamic flux (flow of air along the surfaces of an aircraft) is modified by the aircraft shape. Also the collection efficiency, for a fixed size of cloud droplets, of a wing profile depends upon its form and thickness.

Drop trajectory

Aerodynamic flux

Collection zone

A vulnerability factor: the aerodynamic profile (1)


21

Aerodynamic speed in knots

t

in °

C

1

8

27

Difference in air temperature and temperature of the point of impact (t) in the lower layers of the atmosphere .

Fast aircraft are less vulnerable to ice accretion.

Another vulnerability factor: Aerodynamic speed (2)


22

A turbo reactor in operation can shut down or be destroyed by ingestion of a mass of ice.

Two possible scenarios:• Accumulation on the landing gear while taxiing-out in an area contaminated with frozen snow. The ice lumps will become projectiles on the initial acceleration of the aircraft. • Late use of de-icing or anti-icing equipment during severe icing conditions. The ice that breaks off can fly straight into the engine.

Source Pratt et Whitney Source Pratt et Whitney

Specific vulnerability of the turbo reactor : ingestion of ice during flight


23

This is a diagram which allows us to determine the risk for carburettor icing in function of the temperature and dew point (*).

source Royal Australian Air Force

The carburettors in light aircraft are prone to formation of ice. This icing occurs in the part of the carburettor where the pressure decreases (temperature drops) and where the fuel vaporises (temperature decreases even further). When the icing is important enough, the engine stalls.

The risk is at a maximum in saturated air with temperatures between +5 and +15°C.

Specific vulnerability for light aviation: Formation of ice in a carburettor


24

Light aircraft flying IFR and which do not have effective de-icing equipment are particularly vulnerable in icing conditions.

It is therefore very useful for the pilot to know the lowest flight level (altitude or pressure) where the temperature is below 0°C. (*)

If this sheet of stratocumulus has a temperature below freezing, it would not be a good idea to level out at its level!

Frank Jansen photography

Specific vulnerability for light aviation: flight in clouds under Instrument Flight Rules (IFR)


25

Visual indications for ice accretion (*)

Source ATR

Electronic ice detector (**)

Source ATR

Detection in flight


26

De-icing is the process whereby a “system” removes icing after it has formed on the aircraft.

If the type of icing is not too solid and the intensity of the phenomena is moderate,the pilot can remove the icing by mechanical means.

The advantage of these systems is that they use little energy.

This is why they equip light aircraft and turboprops.

The downside is that these systems can be ineffective in exceptional icing conditions.

Black surface can be deformed by pneumatic

systems

Source ATR

Removal of accretion: de-icing


27

Anti-icing is a system which prevents icing from forming.

The most widely used technique is to heat the elements or surfaces prone to icing.

Advantage: the aircraft will be well protected in almost all icing conditions if they are anticipated.

Disadvantage is that these systems consume a lot of energy. Their use will imply cost penalties.

Source I.A.E

Prevention of icing: anti-icing


28

mea

sure

d li

qu

id w

ater

con

ten

t

median volume diametericing potential

Aeronautical authorities impose exploitation rules during icing conditions. More particularly, certain standards have been defined to certify an aircraft. In this way, the properties of a de-icing system are pointed out.

These standards have been defined on the basis of special studies conducted in the real atmosphere (here CASP II in Canada in 1992). Very extreme situations (red circle), which are rarely encountered are not taken into account.

accretion

severe12 g/cm²/hour

moderate6 g/cm²/hour

light1 g/cm²/hour The conditions which cause the

extreme observations have to be specified.

Certification in icing conditions


29

Experience shows that atmospheric conditions bordering icing conditions have not been evaluated enough.

Actually, these situations correspond to an intense phenomena that is easy to observe or detect and which warrants an immediate and effective reaction from the “aeronautical actor”.

When the atmospheric parameters oscillate around icing conditions or when the conditions are out of the regional or seasonal mean, “traps” ( real atmospheric ambushes) will develop.

Early detection can easily be done by an aeronautical meteorologist.

Marginal weather conditions


30

1) Deliver accurate forecasts within aeronautical range ( 24 hours) , adapted to the local and seasonal context. Determine the first “freezing level” and obtain good scores in forecasting the presence or absence of potential icing conditions on a certain level. These points are operationally important.

2) Achieve a good detection of extreme conditions corresponding to case of observed severe icing. This is an important point for certain categories of public transport aircraft. (commuters)

3) Underline the marginal situations (ambushes) which are bordering on the limit of unpredictable, in order to create a permanent state of vigilance amongst the aeronautical operators. This point is important for air safety in general.

4) Deepen theoretical knowledge about the subject in extreme cases through experiments. This, in order to refine the current standards.

5) Develop training for the “aeronautical actors” on the subject:• To allow them to correctly interpret the information• To familiarize them with the methods and classical scenarios we use• To create awareness about the need of feedback.

Conclusion of the first part : a list with meteorological objectives on icing


31

LFMM SIGMET 3 V

ALID

111145/111545 LFML-

SEV ICE O

BS LFMM FIR

MAIN

LY

ALPS RHONE V

AL CORSICA AND

MAR= AIRM

ET

PIREP

Second part : meteorology and icing

On the topic of contamination by icing, the actions of an aviation forecaster must correspond to the aeronautical objectives which we have defined in part one.


32

The air temperature is an excellent parameter to approach the phenomena. In fact :

• Taking into account the existing relation between air temperature and its capacity to hold water vapour(**), which in a lifting mode can condensate into supercooled droplets, one can state that the icing potential will decrease with the potential temperature, all things being equal.

• Water only freezes when its temperature reaches 0°C or lower. This way one can eliminate all areas with positive temperatures (*).

• Supercooled water cannot exist at temperatures below –40°C. Moreover, because all ice nuclei are active below –35°C, one can consider the icing potential to be nil below –35°C.

• Only a small amount of aerosols make up the ice nuclei between 0°C et –12/-15°C. Condensed water therefore exists mainly in supercooled water. The levels with high icing potential in clouds, are situated in this temperature range.

Prognostic variables of the icing potential in the atmosphere: air temperature (1)


33

The temperature defines for the major part the composition of a cloud.

This useful division has been confirmed by in-flight observations of icing occurrences: few occurrences at temperatures below -20°C (*). Most occurrences were reported at temperatures between -3°C and -7°C (**).



34

Considering the numerical model point of view; temperature has the advantage that it is one of the parameters which is routinely calculated and easyly accessible.

Anyway, its spatial coverage has to be detailed enough in order to exactly isolate the areas with risk for icing and, particularly, the levels with icing potential in the vertical plane. Fine mesh models (mesoscale) are necessary to deliver this necessary precision.

Cross-section across a warm front (distance of the cross 750 km) with a temperature field (blue) and wet-bulb potential temperature in red (*)

0°C level



35

The level(s) of the 0°C-isotherm, more or less determines the lower limit of the icing volumes.

This Météo France product represents the freezing levels in flight levels. The temperatures are plotted if they are sub-zero on surface level.

On a local level and for very short range forecasting, a representative sounding will help to determine the corresponding level(s).

Prognostic variables of the icing potential in the atmosphere: air temperature and the 0°C-isotherm (4)


36

Form the viewpoint of numerical models, relative humidity is also a frequently calculated parameter.

Saturation of air with icing potential can be represented by the relative humidity of the air. (*). If it is described correctly, it is also a parameter which can eliminate areas without icing potential.

Cross section of a warm front (distance of the cross 750 km) with a humidity field (pink) and the wet-bulb potential temperature. (**)

Prognostic variables of the icing potential in the atmosphere: relative humidity (1)


37

It is then quite natural that we associate icing potential to a cross reference temperature/relative humidity (Hu).

In this way, algorithms are based upon a climatological study. Icing potential is at its maximum when you have high relative humidity combined with the most favourable ranges of temperatures.

The determination firstly results in black and white (icing or not). It is more difficult to justify the determination of intensity solely based upon one cross-referenced criterion (*).

Prognostic variables of the icing potential in the atmosphere: relative humidity (2)


38

• Liquid Water Content (supercooled) or LWC expresses the amount of condensed liquid water present in a volume of one cubic meter moist air (generally air inside a cloud) of which the temperature is negative.

• The content of liquid water is a parameter that gives an excellent indication of the icing potential. It is expressed in grams per cubic meter (*).

• It takes a significant value in a saturated environment.

Prognostic variables of the icing potential in the atmosphere: Liquid Water Content

(1)


39

• Stable clouds in layer (St, As, Ns) : from 0,1 to 0,5 g/m3

• Unstable clouds (Cu, Ac, Cb) : from 1 to 5 g/m3 (up to 15 g/m3 in certain unstable clouds in the tropics)

Remark : stratocumulus is a cloud that possess both characteristics, stable on a large scale and unstable on a

smaller scale. Their water content therefore varies a lot.

• Mist and Fog : from 0,1 to 2 g/m3

• Larger amounts are present in precipitation areas.

Amounts of condensed water present in the atmosphere :

Prognostic variables of the icing potential in the atmosphere: Liquid Water Content (2)


40

Stable clouds in layers:

• They have a reduced water content : only 15 to 20% of these clouds contain more than 0,3 g/m3 and only a few of them more than 1 g/m3.• The strongest concentration of liquid water can be found in the temperature range between -5° and -10°C.•In these clouds, which are considered to be homogeneous in the horizontal plane, a big relative variability in water content can be observed.• Water distribution in the vertical plane is irregular. Certain cases have the maximum at the bottom of the cloud, while others have their maximum in the upper part.• There is an interesting correlation (*) between the mean extreme values of water content and the flight distance over which the measurement is done. Between –5°C and –10°C : over 20 km never more than 1 g/m3 , over 100 km never more than 0.7 g/m3 , over 200 km never more than 0.5 g/m3 .



41

Unstable clouds :

• have a liquid water content that frequently surpasses 1 g/m3.• in these clouds, which are in general considered to be heterogeneous, a large variability in water content is observed.

2 g/m3

Measurement of LWC through a Cumulus Congestus field-TCU- at a temperature of –10°C (campaign “ LANDES-FRONTS 84 ”).



42

It is possible to establish an intensity scale solely based upon the liquid water content – LWC (*).

Light icing potential : LWC < 0,6 g/m3

moderate icing potential : 0,6g/m3 < LWC < 1,2 g/m3

severe icing potential : LWC > 1,2 g/m3

(**)



43

LWC is depicted as an output of an experimental high resolution model (Meteo France MESO-NH) in a big convection system (*).

Numerical model output depicting LWC



44

The dimensions of the cloud droplets are in general less than 40 microns.

Measure of Mean Diametric Volume (MDV) of cloud droplets in a cumulus congestus field -TCU- at a temperature of –10°C (campaign “ LANDES-FRONTS 84 ”).

The dimension of cloud droplets determine partially the ice accretion : collection, shape and contaminated zone (*).

The mean dimension of these droplets (**) therefore have to be combined with temperature and liquid water content in order to produce quantitative and qualitative forecasts of ice accretion.

Prognostic variables of the icing potential in the atmosphere: the size of the elements (1)


45

The big droplets (SLD)

Starting from 40 microns (*) a droplet becomes a large droplet (SLD – Supercooled Large Droplet).

Beyond 200 microns they become drizzle and rain drops or supercooled precipitation (FZDZ and FZRA).

Taking into account the major consequences related to a contamination by these big elements of supercooled water, we have to discriminate between

SLD scenarios and others (**).

Prognostic variables of the icing potential in the atmosphere: the size of the elements (2)


46

Sometimes vertical velocity is used as a complementary parameter to discriminate icing conditions, especially when you do not have a model prediction of LWC at your disposal.

The idea is that the water present in the clouds originates from the base of the atmosphere and that it has been brought there by vertical movements. All things being equal, the more intense these currents are, the more important the upward flux of water is (*).

The advantage is that numerical models predict vertical velocity.

Experience shows however that a good discrimination (**) of icing layers require a mesoscale description, which is a scale at which the operational models are not always very efficient.

Prognostic variables of the icing potential in the atmosphere: Vertical Velocity (1)


47

The vertical velocity field is here combined with an icing index T/Hu. The icing index (pink isolines) is “filtered” by the vertical velocity with values below –0.2 hPa/s in the red areas.

A cold front reaches Europe. In the north-west of Spain, significant icing conditions were reported. At the same time, elevated icing index values (8 to 9 on a scale of 10) indicate a high risk of icing (*).

Prognostic variables of the icing potential in the atmosphere: Vertical Velocity (2)


48

Taking into account the importance of the stable or unstable nature of a cloud with respect to its LWC, it is clear that anticipating this characteristic is part of the forecasting icing conditions process when one does not have a numerical model output of LWC at its disposal (*).

Vertical instability

Wind Shear

We will see that wind shear can favour conditions with presence of SLD (**).

Prognostic variables of the icing potential in the atmosphere: Vertical Instability – Wind shear


49

CLASSIFICATION

Class I Light Icing

Class II Moderate Icing

Class III Severe Icing

Mean quantity of supercooled water

< 0,6 g/m3

0,6 - 1,2 g/m3

> 1,2 g/m3

Mean diameter of droplets or drops

< 50 µm

> 50 µm

Nature of clouds and hydrometeors

Stable ---------------------------------- Unstable

Very unstable Freezing rain and

drizzle

Icing types Hoar Frost Rime Ice

Rime ice Mixed ice Clear ice

Mixed ice Clear ice Rain Ice

Conclusion about Icing Potential

Situations with severe icing are equivalent to a high LWC or high amount of large droplets or drops (lots of SLD).


50

An unstable scenario (convective)

This corresponds to the possibility of an intermittent, but important, vertical convective transport of water vapour from the lower layers towards a level where the conditions favour the condensation of vapour into liquid supercooled water.

This can be observed in all situations leading to the presence of low or medium level unstable clouds.

The vertical temperature profile serves to determine the freezing levels (down to -20°C) and in some cases the icing potential, which is strongest at the base of the negative layer within the cloud (*).

Favourable scenarios for icing : convective (1)


51

This sketch illustrates a scenario where lots of ice nuclei are present at the base of the cloud. These numerous nuclei multiply the number of smaller elements (light ratio of SLD)

a

This sketch illustrates a scenario with the presence of large supercooled elements in the “heart” of the cloud (*). Large drops become supercooled (high ratio of SLD)

b



52

This profile, measured in the northwest of Spain, indicates that there is a lot of available energy for convection (high CAPE). The sounding was made ahead of a cold front. It indicates possible development of convective clouds (deep convection). These clouds represent a strong icing potential even at high levels (close to -20°C). These conditions were confirmed by two in-flight observations of ice accretion in this region (*).



53

A turbulent scenario

This scenario corresponds to a continuous turbulent vertical transport of water vapour from the lower levels towards a level where the temperature favours condensation into supercooled water, and where a temperature inversion has a blocking effect.

The vertical temperature profile is used to determine the icing levels (down to -10°C) and sometimes the icing potential, which is in general stronger in the top of the cloud (**).

The inversion caps the vertical transport, concentrates the turbulence and favours high water content in the cloud top as well as the presence of SLD (*).

Favourable scenarios for icing : turbulent (4)


54

Or how you can find a strong icing potential at the top of a stratocumulus layer.

t

Z

0°Ca 0°C

t

b

Few crystals and lots of SLD

Lots of crystals and few SLD



55

1000 hPa

850 hPa

700 hPa

stratocumulus SC

Ingredients :- turbulent saturated layer with vertical wind shear- strong subsidence and vertical capping due to inversion-temperature range between –3°C et –7°C- IR radiation during the night which reinforces the cooling at the top of the stratocumulus cloud.



56

A frontal scenario

This corresponds to the possibility of finding supercooled droplets (SLD, FZDZ and FZRA) in the “sheared” atmosphere (shear of the wind) within a frontal region.

A “sheared” atmosphere is often found in the vicinity of a warm front with a warm layer (with positive temperatures) trapped between, two layers with negative temperatures (*).

Favourable scenarios for icing : frontal (7)


57

Areas with strong icing potential are usually more extensive ahead of the warm front, in the heart of occlusions and in the vicinity of anabatic cold fronts.

(*)

Favourable scenarios for icing : frontal (8)


58

As a result of the sum of different changes, the icing potential generally increases above mountains (*).

Orographic scenario

The height of the 0°C isotherm rises with unstable air and lowers with stable air.

0°C

Favourable scenarios for icing : orographic (9)


59

O°C

Stable clouds as a result of orographic lifting (*) have sometimes high concentrations of supercooled liquid water.

Stable Layer

Orographic scenario



60

Orographic scenario



61

Freezing Fog

If you expect a rapid cooling of the fog layer, you should anticipate a worsening of the icing conditions.

Favourable scenarios for icing : freezing fog (12)


62

In the near Infrared channel (1.6 microns)

Detects liquid water in clouds (reflectance factor)

On this daytime picture, you visualise in the A region relatively low liquid water clouds. A

B

High level ice clouds in the region B are less apparent.

Satellite observation (1)


63

In the middle Infrared channel (3.8/3.9 microns)

Detects low clouds and fog (emissivity)

This daytime picture has a low contrast. But, during the night, without sun radiation, the thermal radiation is well depicted.



64

In the thermal Infrared channel (8.7 microns)

Detects composition of the clouds (emissivity) : ice or water

B

Thin cirrus are more apparent (B). As they prevent from a good observation of lower clouds, they are easily localised.



65

In the thermal Infrared channel (10.8 microns)

Detects cloud top temperatures or temperature of earth’s surface in the absence of clouds

This is the usual infrared channel. It can be crossed with another IR channel (12 microns) in the way to reduce the effect of atmospheric absorption.



66

Water Cloud(grey)

MODIS Image over Great Lakes, 1 Mar 2002, 2.52 UTC results for cloud phase of the (ch. 8.6, 10.8 and 12.0 m) MODIS overpass

Clear Land(green)

Clear Sea(blue)

Ice Cloud(red)

Mixed Cloud(white)

Determination of the cloud phase using satellite observations



67

Composite images: firstly they can be used to select the cloud regions according to operational cloud classification. Then, when combined with the SEVIRI METEOSAT 3.9 and 10.8 micron channel (thermal infrared), clouds with icing potential are isolated if they are not underneath very cold high cloud (*).

This is an experimental picture by Météo France where the red areas are clouds with liquid water in the top.



68

The icing layer in the vicinity of fronts have often a structure which can be compared to the precipitation echoes (areas with updraughts where precipitations will occur).In image a, elongated orange areas of 15 to 30 km wide and less than 150km long are observed. In image b, a yellow/red structure of 2 to 4 km wide and 15 to 30 km long is depicted.

Radar observation (1)


69

If this is present in the vicinity of a front and above a layer with negative temperatures this is favourable for a freezing rain or freezing drizzle scenario (*).

If you have radar images in the horizontal plane or in a cross-section mode, they can be used to detect or confirm icing conditions.

The bright band corresponds to a volume with strong reflectance in the melting layer of the precipitation.

Radar observation (2)


70

Three main approaches for forecasting icing conditions can be identified:

1. You extrapolate in space and time icing observations.

2. You document icing situations with the help of observational data and standard meteorological forecasts. Then you will use this information to identify scenarios known to be favourable for icing.

3. You take advantage of an icing index produced by a numerical model.

Forecasting techniques (1)


71

If you have a reliable source of icing observations (accretions) coming from airspace users, they can be used to depict the phenomenon and to anticipate it by appropriate extrapolation in time and space.

When you have frequent observations (PIREP or SPECIAL AIREP) nicely spaced, you can consider an operational follow-up of the phenomena (in a very short range forecast mode) (*)

In any case, this approach is often biased by the difference of the observations.

This has the advantage that the users are implicated in the forecasting process which can help in detecting the flaws in the other forecasting techniques. This can also be used to document extreme situations.

Forecasting techniques: extrapolation of observational icing data (2)


72

This picture displays the in-flight icing observations coming from PIREPs over the US. They have been superimposed on a numerical model output of SLD icing potential.(*)

Forecasting techniques: extrapolation of observational icing data (3)


73

With good knowledge of the meteorological situations favourable for icing conditions, each time one finds such a scenario (*) in the observational data analysis (**) or in the numerical model outputs analysis, one has to inform the users.

This can be a good method to identify extreme situations.

Drawback: it is a difficult method to put in place over large areas and needs a costly local expertise.

Forecasting techniques: looking for icing scenarios (4)


74

If you have at your disposal fine mesh numerical models (mesoscale), one could consider an output of an icing index (*).

This method has to be supported by a regular evaluation through comparison with observational data, and has to cover all the classic scenarios.

This method has the advantage that it is easy to put into operation for large areas.

The quality of the forecast is correlated to the model’s ability to reproduce complex processes on a small scale: realistic microphysics, sophisticated but robust schemes (**).

Forecasting techniques: using a numerical icing index (5)


75

cloud droplets

condensation-evaporation

water vapour

ice crystals

snowgraupeln

rain drops

auto-conversion,accretion

Bergeron-Findeisen process,homogeneous nucleation

riming, dry and wet growth

melting

This is an explicit scheme (*) which is at work in the MESO NH model (**).Cloud LWC is an output of such a process.

Forecasting techniques: using a numerical model output (6)


76

To forecast icing conditions, research is ongoing to automatically combine the model output indices with combined observational data (satellite, radar, PIREP).

ADWICE (Germany) combines the output of the LM model with conventional observations (SYNOP, METAR) and radar, but does not use satellite data.

CIP (US) is an index of icing severity that combines de output of the RUC model with all possible icing observations, PIREP’s and electrical detection included.

SIGMA (France) combines the output of the Aladin model (air and surface temperature, humidity) with satellite and radar observations (*).

Forecasting techniques: crossing observations with model outputs (7)


77

FZDZ high risk under warm clouds tops

FZDZ high risk under cold clouds topsFZRA light risk under cold clouds tops

high risk of icing under warm clouds topslight risk of icing under warm clouds tops

light risk of icing under cold clouds tops

high risk of icing under cold clouds tops

FZDZ light risk under cold clouds tops

Example of a SIGMA image (Météo France) (*)

The use of these combined products needs a good knowledge of the resources and algorithms and a validation for local use (**) (for example in a mountainous region).

Forecasting techniques: crossing observations with model outputs (8)


78

In general, it seems that the latest automated products have the ability to distinguish (*) the icing cases from de non icing ones (risk/no risk), but the results are marginal when identifying the situations with severe icing. Moreover, convective situations are usually baldy captured by numerical models.

It is therefore that when a local study (**) is added, the scores of detecting severe icing will improve to the satisfaction of all aeronautical users (***).

But the constantly improving remote sensing techniques and numerical models leave us with a good margin for improvement.

Forecasting techniques: towards a conclusion (9)


79

Forecasting icing conditions falls in the common and routine anticipation of significant weather for the aviation world in the same way as for instance the forecast of atmospheric turbulence.

In any case, regional or local climatology sometimes gives more importance to this forecast. :-more impact outside the tropical regions-more impact in wintry conditions in the mid latitudes areas

Each Meteorological Service has to adapt its efforts (*) to the local situation and the needs expressed by the users of the forecasts.

Putting in place icing conditions forecasts (1)


80

The quality of the service depends for a large part upon the means available for bringing forecast into play.

Beforehand :- An observational system (of mesoscale and aerological scale) : including multispectral satellite images and radar images but also conventional surface and upper air observations

Afterwards:-An adapted numerical model (mesoscale) with explicit schemes to produce prognostic icing parameters or an icing index -a process to combine all targeted observational data.- a local expertise(*)- a high performing transmission system

Putting in place icing conditions forecasts (2)


81

Aeronautical messages SIGMET et AIRMET (*) are the standard formats.

They are produced by the meteorologists in charge of the Meteorological Watch Offices (MWO) for the flight information regions (FIR and UIR) and have to be in accordance with ICAO recommendations.

A SIGMET message can describe forecasted or observed severe icing outside convective clouds by using the code SEV ICE . (FZRA or FZDZ) is added when the phenomenon is caused by freezing rain or drizzle.

Example : LFBB SIGMET 3 VALID 070500/070900 LFBD-LFBB BORDEAUX FIR SEV ICE (FZRA) OBS FCST BTN GND AND FL060 LINED AHEAD WARM FRONT W OF LFBB FIR 75 NM WIDTH MOV E 15KT NC

A transmission chain of adapted information: alphanumerical formats (1)


82

A GAMET (**) has the possibility to indicate moderate icing outside convective clouds in section 1 for the lower levels.

An AIRMET message can describe (in the same format as a SIGMET) observed or forecasted moderate icing (which has not been mentioned in the SIGMET) in the lower levels (*) outside convective clouds by using the code MOD ICE.

It is also recommended that the aerodrome warnings also include following phenomena which imply a risk of contamination by ice: snow, freezing precipitation, hoar frost, clear ice, rime ice, frost.

A transmission chain of adapted information: alphanumerical formats (2)


83

Significant weather charts (SIGWX) explicitly indicate the moderate or severe icing potential (*) (**)

Example in chart format SIGWX EURWAFC LONDON

Moderate icing potential in clouds between, level XXX (here FL100 the first to be describe in this chart) and FL170.

A transmission chain of adapted information: graphical formats (3)


84

In-flight observations of aircraft can include information on observed ice accretion (*).

Collected by the meteorological watch office, they can be centralised in the form of PIREP’s in order to validate the forecasts.

For safety raisons, it is highly desirable that in the situations with severe icing, the feedback of information by the MWO, is fast and up to date (**)

In order to perform a quality control on the forecasts, it is desirable to set up a light, regional level, procedure for collecting observational data (***).

A transmission chain of adapted information: Observations of ice accretion (4)


85

These are a few attitudes (*) which could degrade our work:

On the meteorologist side,- a routine approach to the forecast- little knowledge of the aeronautical world- being isolated from the users

On the user of the forecast side,- simplification of the phenomena- loss of vigilance at the start of the winter season or because the phenomena is rare- little knowledge of the meteorological world and the limits of icing forecasts- too much or too little confidence in the forecasts- not transmitting systematically the own observation of the phenomena on a routine base.

Awareness of the icing phenomena (1)


86

These are a few attitudes that can improve or work:

On the meteorologist side,

- recognise the marginal situation (on the edge of…) that are often forgotten and which can lead to « booby trap » situations- be reactive to observations of severe icing (*)- inform and mobilise the users on the progress in concepts and the evolution in forecasting techniques of the phenomena (**).

On the user of the forecast side, - use the ability to observe the phenomena and transmit the obtained information (***)- inform yourself regularly on the progress of the concepts and the evolution in forecasting techniques for predicting icing

Awareness of the icing phenomena (2)


87

-towards a better understanding of physical processes in cloudy environments.

-sophistication of the cloud schemes: explicit representation of the physical processes in the mesoscale modelling

-in-flight observation campaigns in order to sample the regional situation

- Objective and active standardization of the response of aircraft to the icing conditions. (*)

A few points on where to go with research and development


88

In english :A small complete book : Aircraft Icing HandbookGAP booklet, Safety Education and Publishing UnitCivil Aviation Authority P O Box 31 441, Lower Hutt, New Zealand

In french :Un ouvrage plus général : Météorologie AéronautiqueEcole Nationale de l’Aviation Civile7 Avenue Edouard Belin B.P. 400531055 Toulouse Cedex 4, France

Common International training standards : •WMO No 49, Technical regulations vol II, Meteorological Service for International Air Navigation •WMO No 258, Guidelines for the education and training of personnel in meteorology and operational hydrology•WMO No 732, Guide to practices for meteorological offices serving aviation•ICAO Annex 3, Meteorological Service for International Air Navigation

Bibliography (*)


89

The COMET program (US) :-On icing in general http://meted.ucar.edu/icing/pcu6/-On forecasting icing by using observationshttp://meted.ucar.edu/icing/pcu62/pcu622/ -On forecasting icing using profileshttp://meted.ucar.edu/icing/pcu62/pcu621/

Environment Canada :Icing and aeronautical meteorology http://www.msc-smc.ec.gc.ca/armp/icing_e.cfm

Example of operational data on icing (US)http://adds.aviationweather.noaa.gov/icing/

Aviation accident reports (France)http://www.bea-fr.org/francais/rapports/rap.htm

A few internet resources on the subject (*)

http://meted.ucar.edu/icing/pcu6/





http://meted.ucar.edu/icing/pcu62/pcu622/










http://www.msc-smc.ec.gc.ca/armp/icing_e.cfm












http://adds.aviationweather.noaa.gov/icing/







http://www.bea-fr.org/francais/rapports/rap.htm









90

-Contamination, contaminant, ice-Accretion (of ice), hoar frost, rime ice, clear ice, mixed ice-Icing (light, moderate, severe)-Icing conditions; icing levels-Icing potential (light, moderate, severe)-Carburettor icing-De-icing, anti-icing-Temperature, 0°C-isotherm-Liquid Water Content (supercooled)-Large droplets (SLD)-Vertical wind shear-Freezing fog-Satellite observation (near infrared, middle, thermal)-Radar observation, bright band-Icing indices-SIGMET, AIRMET, GAMET, SIGWX charts, PIREP

List of key words (*)


91

This icing case study has been established with resources which are available within Météo France.

Ideally, other icing case studies, provided by other meteorological services, could widen this presentation of forecast techniques.

Of course, the forecaster in charge of aviation will be able to apply these forecast techniques only when equivalent resources are available.

However, the meteorologist can use these techniques to adapt/develop his own local methods.

Case study (1)


92

Observation 2005 April 2 :Severe icing has been observed during a flight in central Spain, during a period of only one hour and thirty minutes (between 0945 and 1115 UTC) at flight levels between FL100 and FL130 (700 and 600 hPa) with temperatures under –6°C and above –11°C.

This case study should be viewed as a routine forecast for icing conditions.

Case study (2)


93

These icing observations are situated behind a stormy front which was moving towards the north-east.Convection was growing in this cyclonic region.This convection continued to intensify during the morning, as a result of diurnal evolution. On 2 April, in this position, the sun rises at 0600 UTC and at the time of the icing observations, more than 3 hours of warming within the low troposphere had been going on. The geographic environment is a mix of plateaus (700-900 metres altitude) and mountain ranges (2000 metres altitude).

Case study (3) : synoptic situation


94

The airmass, within which the convection developed, is represented by the radio-sounding of Santander at 1200 UTC. Santander is situated about 150 Nm downwind of the icing observation region.

A more precise analysis indicates a CU/TCU/CB development, with bases under 800 hPa and tops growing up to 700 or 600 hPa for CU/TCU and at least 350 hPa for the CB’s.

The CAPE index is rather high (592 J/KG) with sufficient moisture. This aerologic situation favours the growing of deep convection.

Case study (4) : air mass observations


95

Inside the icing observation area, the IR 0945 UTC satellite picture (European satellite Meteosat 8) presents only small cloud systems which are isolated and which probably correspond to convective cells.

In the icing observation area, the temperature of the cloud pixels are close to –2°C.

Case study (5) : satellite observations


96

This picture is a result from a process which uses different satellite pictures (Meteosat 8) at the same time (0945 UTC) and represents a “cloud classification” picture.

Inside the icing observation area, there are a great variety and fragmentation of colours according to the convective situation.

Pixels show very low clouds (probably ST after rain showers) with semi transparent clouds on top (probably Ci CbGen) which spread out under influence of the south-easterly wind in the upper troposphere.



97

However, at least some pixels indicate potential icing cloud at the exact time and position of the icing observation.

Le 02/04/2005 à 0945TU

This picture is a different combination of satellite pictures (Meteosat 8) at the same time (0945 UTC) and represents a composite picture where potential icing clouds appear in red.

Inside the icing observation area, conditions appear very variable. A few pixels show potential icing clouds. A few pixels show too warm or too cold cloud tops and other pixels are unspecified.



98

Radar imagery at 0945 UTC presents a few moderate but isolated echoes which correspond to the presence of deep convection inside the area with icing observation.

Case study (8) : radar observations


99

The previous data leads us to a convective scenario

This indicates the occurrence of potential intermittent icing. The intensity can be moderate to severe in the CB.

Inside the CBs, the levels of 0°C (780 hPa) and – 20°C (520 hPa) isotherm indicate that the icing flight levels range from FL070 to FL170.

When you consider the level of the 0°C isotherm (far above the cloud base), SLD is likely to be encountered inside these clouds during the developing stage.

Case study (9)


100

Picture SIGMA at 0945 UTC, which has been established using the fine mesh numerical model ALADIN (9 hours forecast) completed with observations. 3 colours for 3 icing risk thresholds (icing potential). -yellow : low risk-orange : moderate to high risk-red : high risk associated with precipitations

Inside the area with icing observations, SIGMA is not effective

This analysis which has been supported with observations and linked to a specific scenario, could be completed by using numerical model outputs. Below is the output of the Météo France SIGMA system .

Case study (10)


101

The icing index of the SIGMA system is calculated with both temperature and relative humidity values which are extracted from the model. Thresholds are 0°C and –15°C with humidity above 80%. The forecasted humidity at 0900 UTC inside the icing observation area was only 75%!

This convective situation is difficult to capture, as long as you use a model which is not capable to give an explicit description of convection.

On the other hand, in the north-east part of the icing observation area, an area where you find a dynamic ascent linked to the frontal zone, the SIGMA picture is much more accurate in detecting the risk for icing. Red indicates a criterion of high risk, where the cloud tops are relatively warm. Supercooled droplets will likely be encountered. To determine the intensity of icing, you will have to define the risk of finding SLD or freezing precipitation.

Case study (11)


102

This situation, which has to be incorporated in a routine forecast of the icing hazard, needs to be anticipated at least 6 to 9 hours before its occurrence. This is the normal lead time for a significant weather chart forecast.

To succeed in this challenge, the meteorologist in charge of aviation generally has at his disposal : • a synoptic weather analysis and forecast • a serial of prognostic variables which are outputs of a numerical model • observation data which could be extrapolated

He will add sometimes a numerical valuation of the risk in the form of an icing index or something similar.

You will notice that in scenarios like this convective situation, a careful examination of a radio sounding is very effective to describe the nature and the vertical extension of the risk.

On the other hand, you will have to downgrade the use of a numerical model as long as the description of the convection is not explicit.

Case study (12)


103

Significant weather forecast 2005 April 2 0900 UTC

The region where icing has been observed is included in an area of strong convection. CBs and thunderstorms are forecasted. A moderate icing symbol is printed.

Inside the reference frame of this chart, you read the mention : “CBs or thunderstorms mention implies moderate to severe icing”. (probability to encounter large super cooled elements)

Case study (13) : significant weather chart

Documents

Freezing contamination : aircraft icing