Print Click here to print this project

i-WAKE2: island-induced WAKE dynamics

Start date: 25-08-2010 - End date: 05-09-2010

Status: Confirmed

Open to sharing: Yes

Confidential: No

Transnational Access: Yes

Open to training: Yes

Grounded / Maintenance: No


Aircraft name: ATR42 - SAFIRE

Airport: Madeira atmospheric wakes illustrated well the VKVS examples in the classic scientific literature (e.g. Scorer, 1986; refer to VKVS image attached). Due to its location (in the NE Atlantic Ocean) and considering: (i) the size and orientation of the island’s mountains; (ii) the predominance and strength of the NE trade winds; VKVS frequently form leeward of Madeira and are very visible from above, reaching as far as 600km offshore. Furthermore, the phenomenon is very predictable using numerical forecasting systems such as MM5. In fact, VKVS are frequently predicted (every week) with our operational weather forecasting system ( => Operational Products => Meteo =>Weather Forecast). Madeira Archipelago oceanic region is also an area deprived of oceanographic / meteorological observations and the current project, in particular the numerical studies, will substantially benefit from airborne data. Araujo et al (2010) analysis of QuikSCAT data also showed that wind shear zones, during wind wake episodes, are regions of intense oceanic eddy activity. SAR data analysis confirms the concurrence of these wakes leeward of Madeira Island. Furthermore, several team members are currently involved in the European Space Agency (ESA) project (6248) which allows for scheduling concurrent SAR and airborne data collection (Near-Real Time). This is a rare opportunity since programming of SAR missions is only possible in the scope of pre-approved (peer-reviewed) ESA projects. SAR sensor is often turned off over oceanic regions. The main objective of the current ESA project is the study of the origin and fate of Madeira Island induced features (refer to Araujo et al., 2010 attached).

Project description

Project theme: air-sea interaction

Science context: In geophysics, "island wakes" is a term typically used to refer to atmospheric circulations induced by mountainous islands (atmospheric wakes) as well as ocean effects induced by the islands' bathymetry (ocean wakes). In turn, ocean wakes,can be grouped into two main categories (i) wakes induced by atmospheric phenomena (wind wakes a.k.a. ‘warm wakes’), and (ii) wakes induced by oceanic phenomena. Both atmospheric and oceanic wakes have been the subject of many studies by the scientific community. Attached please find a reference list of some wake studies (REFERENCE_LIST ATTACHED). Reynolds number theory predict that island wakes may vary from attached vortices to fully turbulent wakes resembling Von Karman Vortex Streets (VKVS). Froude number laboratory studies lead to the development of the concept of the “diving streamlines” in stable flows, whereby by the relation between atmospheric stratification and the height of the mountain lead (or not) to the generation of VKVS. Currently, Madeira Island wakes are being studied using numeric and laboratory models, satellite remote sensing, and in situ observations (Araújo_etal_2010). However, none of the available sampling means can give us a synoptic (i.e., quasi-simultaneous) high-resolution view of atmospheric and oceanic wakes; therefore, we hereby propose to carry out airborne observations of atmospheric wakes (high-altitude) followed by airborne observations of oceanic wakes (low-altitude), by carrying out measurements in a region where both types of wakes frequently occur. Our proposed target region is the lee side of Madeira Island (33N;17W).

Measurements to be made by aircraft: Scientific objectives: Multidisciplinary and multiplatform approach to: (i) characterize spatial and temporal characteristics of Madeira Island atmospheric wakes; (ii) measure sea surface signatures of leeward oceanic eddies; (iii) study the contribution wind stress forcing to the generation of oceanic vorticity; (iv) study the effects of solar radiation and cloud coverage feedback mechanisms to the formation of oceanic warm wakes (i.e. ‘warm pools’). Proposed work: (i) high-altitude transects across the wakes; leeward of the Madeira Island (see EXAMPLE of a proposed sampling plan map-view); (iii) dropsondes above the ocean; upstream of the island, in transit from Porto Santo to the leeward of the Madeira, to obtain necessary information on the upstream atmospheric structure for the wake characterization, and on the leeward side to characterize the vertical structure of the wake region (; (iv) low-altitude transect flights to capture the sea surface signal of oceanic eddies and warm wakes. The exact flight pattern including spacing between wake transects will be determined based on the real-time numerical predictions (MM5) and satellite observations. The flight pattern and the sampling grid should adequately resolve the horizontal structure of wind stress and SST. Considering a VKVS scenario, equivalent to the ones documented by Young & Zawislak (2005), and considering the fact that the two main mountain ranges in Madeira reach between 1265m and 1898m altitude, using an MM5 setup, the leeward along-distance between atmospheric eddies (a) might vary between 98km and 134km; considering Madeira’s island size, leeward atmospheric eddies are expected to have diameters (h) of ~50km. Thus campaigns aiming at sampling the full extent of the wake (cyclonic + anticyclonic shedding) should a.k.a. “far-field” study are expected to cover areas of ~150x50km; whereas eddy targeted campaigns (“near-field”) are expected to sample an area of ~50x50km. Considering the shedding period (f=Ue/a), where Ue is the eddy propagating speed, f=5.92hours (for 1265m mountain case); and f=8.95 hours (for the 1898m mountain case). Therefore, airborne campaigns need to take 5hrs or less to complete in order to maintain a synoptic sampling plan. Given the climatology, it is reasonable to expect that several missions will be flown during the requested period. Thus, we envision 2-3 missions being focused on the ‘far-field’, and 2-3 missions on the characterization of individual vortices (‘near-field). The’ transect sizes and distances between individual transects will be selected to correspond to the spatial scale of the processes to be sampled. To provide adequate flexibility in the flight planning, SAFIRE is currently applying for permissions to carry out these flights in a general (wider) region over the NE Atlantic (see general region map attached). According to information obtained from the pilots, the final flight plans (within the general region) can be provided up to 24 hr in advance of the flight. To facilitate easier flight planning and selection of optimal flight tracks, intensive efforts will be made ahead of the field campaign to compile numerical prediction products and satellite observations in ‘Google Earth. It is expected that general far-field research missions will be composed of longer and more separated flight legs, whereas near-field research missions will be have of finer sampling grids. For instance, to cover far-field sampling area, distance between legs are expected to be ~10km apart whereas near-field campaigns can afford smaller grid spacing and thus higher spatial resolution (~5km apart). 5 Km grid resolution can adequate characterize a 50km eddy system as well as a 10km grid spacing can characterize a 150km wake phenomena. Anticipated output: It is expected that we will be able to produce maps of the wind direction, wind speed as well as the wind stress curl variability leeward of the island, for both the high- and low-altitude flights. From the wind curl we will be able to calculate the Ekman pumping velocities, which are though to be the ocean eddy response to atmospheric forcing (Barton et al., 2000). From the wind speed, humidity, upwelling and downwelling solar radiation, and air and sea surface temperatures we plan to derive the air-sea heat fluxes (latent, short wave, long wave and sensible heat), which are most likely responsible for the oceanic warm wake generation and dynamics. Maps of sea surface temperature from the low-altitude flights will help study the relationship between oceanic eddies and atmospheric eddies as well as the occurrences and spatial extent of warm wakes. Cyclonic eddies have a low sea-surface temperature signature (upwelling), whereas anticyclonic eddies have a higher sea-surface temperature, relative to the surrounding waters. Wind-induced warm wakes form a "pool" of warm water at the ocean's surface; thus leeward dimension of the warm pool is expected to be a function of the exposed (cloud free) regions (downwelling radiation), previously revealed by the satellite dataset (Araujo et al.,2010). We plan on using the in situ (land meteo stations) and dropsonde data from the wake survey flights in campaigns to help validate the numerical model simulations. There is also a possibility of combing the proposed aircraft mission’s campaigns with concurrent oceanic surveys. Airborne campaigns could help steer ocean measurements to a particular oceanic eddy or warm wake region. As shown in Caldeira and Marchesiello (2002), shallow diurnal oceanic thermoclines often form in warm wake regions, leeward of islands. Vertical characterization of oceanic thermoclines requires in situ observations using a local research vessel. It is also expected that the EUFAR airborne efforts in this project will complement another project (GEOMAD) just by adding a two flight transect over land to measure the geoid over Madeira with a 10cm accuracy. Dropsondes are to be used only over the ocean to characterize the atmospheric stratification windward and leeward of the island’s mountains. Successive research missions (for example, morning and afternoon) can be used to study the spatial and temporal variability of island wakes, as well as to increase the probability of capturing oceanic features. Day and Night time campaigns can help understand the diurnal variability of upwelling /downwelling solar radiation and its oceanic (feedback) response.

Season: Summer months 2010 (August/September). Island wakes are very common leeward of Madeira Island. Predominantly northerly incoming winds interact with the island mountain’s inducing the formation of leeward vorticity on a daily / weekly basis. In turn, an asymmetric atmospheric vorticity fields uncovers the sea surface to allow for the formation of a ocean warm pool. The stay of the aircraft for two weeks in the archipelago would enable the study of the temporal and spatial evolution of some of these features.

Weather constraints: Mean wind speed in the Madeira Island region varies between 2-15 m/s (1995-2005); most often the winds are from the north quadrant (N; NE; NW). Observed and calculated (MM5) vertical profiles for Madeira, show a two layer structure with a strong inversion layer defined at ~1000m, well below the mountain’ top. Under these conditions, the atmospheric wake is expected to be present most of the time but its life span will depend on the wind speed and the atmospheric stability. Northerly winds are the most appropriate for the formation of VKVS although incoming south winds have also been observed to form a north wake. North atmospheric wakes only last 2-3 days, whereas south wakes can last for several weeks, due to the predominant trade wind pattern. We currently run a weather forecasting system; which allows the accurate prediction of the local weather including the formation of leeward atmospheric wakes with a 3-days notice (refer to NorthWake_SAR&MM5 as a forecasting example). Summer conditions are often clear over the ocean (no fog) and best suited for airborne and oceanic campaigns. Occasionally, the Madeira atmospheric VKVS consists of large wake extending 400-600 km offshore. Nevertheless, on a daily / frequent basis the Madeira mountains interact with the North incoming winds forming VKVS at smaller scales (extending 150km offshore). Atmospheric vortices leeward of Madeira, in their early formation stages often measure 15-20 km in diameter, which are on the edge of satellite detection (e.g. QuikSCAT). Furthermore, QuikSCAT dataset is masked by land, limiting its use in detailing shear zones next to the island’s flank (see examples of Araújo et al 2010). As part of a Master thesis effort (2007) we validate the atmospheric forecasting model (MM5) for Madeira region which allows to forecast the wind patterns with an acceptable accuracy at different model resolutions (R=0.979; RMSE = 3,233) (; English summary attached to this proposal: MM5_Madeira). Therefore we are confident that we can predict these daily-episodes of atmospheric wakes leeward of Madeira. Numerical studies using MM5 also suggest that the atmospheric (vertical) boundary layer in intense north-wind episodes is often located bellow the altitude of the highest mountains. Therefore, lower clouds are often located between 1200-1500m. Trips to the top of the Madeira Mountains can often take tourists above this cloud/boundary layer, during summer months for spectacular views. Upwind of the island there are often strato-cumulus clouds generated by wind blockage. In general, flight conditions are expected to be turbulent near the island flanks and directly leeward of mountains and calm in the remaining of the leeward (warm wake) region, except at the lowest levels where one expects boundary-layer turbulence.

Time constraints: Satellites of interest are NOAA-15,16,17,18, MODIS and ENVISAT, JASON-1/2, as well as TOPEX- Poseidon; however, since they are polar-orbiting satellites they have more than one over-pass per day, therefore they do not pose any time constrains for concurrent data collection. Nevertheless, passive infrared sensors flying onboard of NOAA and MODIS satellites are limited by cloud cover and altimetry passive radar sensors from Topex-Poseidon and Jason 1/2 have limited spatial resolution (10x10km / pixel). We hope to combine the use data from active radar sensors, flying onboard of ERS-1/2 Synthetic Aperture Radar (SAR), and ENVISAT-ASAR , with our MM5 weather forecasting numerical toolkit, to better predict and study the temporal and spatial scope of atmospheric wakes. NRT SAR data can also be used to help plan flight campaigns (see for example: NorthWake_SAR&MM5). Notwithstanding all passive sensors limitations, it is preferable to carry out the airborne campaigns during (typical) summer months, in order to maximize the availability of a multiplatform dataset. It is desirable that the EUFAR airborne campaigns, leeward of Madeira are scheduled for a minimum period of two weeks. This would (i) maximize the chance for sampling a variety of several wake scenarios; and (ii) increase the probability of achieving concurrent measures for atmospheric and oceanic phenomena i.e. eddies and warm wakes.

Flights (number and patterns): It would be important to sample the atmospheric and oceanic wakes in several occasions using airborne technology in order to study their temporal and spatial dynamics. Atmospheric VKVS are better sampled with higher altitude flights (above the clouds), including the use of dropsondes; whereas oceanic wakes are better sampled under the clouds i.e. low altitude flights over the ocean. Airborne sampling transects are expected to be completed within 5 hours. High altitude flights (offshore path) are to be followed by low-altitude flights over the same region i.e. return flight (inshore path). Most data collection is to take place over the Atlantic Ocean not over land. During the 10 days it is expected to have at least one campaign per day. Occasionally, and provided that international regulations are meet, daily flight can be followed by the same night flight path, in order to study the feedback mechanism onto the ocean proposed by Hafner&Xie (2003). The possibility of the aircraft to be stationary in one of the islands (Porto Santo) for the whole duration of the project would maximize the quantity and quality of the data collected. In addition to sample at least two flight levels in each campaign (61m and 2000m), we also plan to carry out an aircraft vertical profiling, whereby the whole aircraft would do a spiral sounding (up and/or down and up) to obtain information on the upstream/downstream atmospheric structure. For the shorter near-field campaigns, such as those targeting the characterization of a single eddy region it is possible to fly the aircraft at several altitudes (61; 500; 2000m). Considering a cruise speed of 100m/s and a 5hrs flight can cover a maximum 1800km. Undulated (ascending/descending) aircraft behavior is expected to cost -14% of flight path, thus conditioning the maximum flight path to ~1500km in 5hrs. Attached to this proposal we include several examples of far-field and near-field campaigns to study south and north wake episodes, in addition to proposing a general flight campaign. Nevertheless, we hope to use ‘Google Earth’ to combine MM5 forecasts with satellite data and possibly include online in situ (meteorological data available online) data to plan daily flight campaigns. Daily discussions amongst pilots, scientists and technical team members, in tune with local authorities, will continuously improve campaign efforts. “Quick looks” of previous day dataset, will also generate valuable cumulative knowledge to improve the efficiency of the daily campaigns.

Instruments: - 2 Pyrometers Kipp&Zonen CMP22 (visible) and 2 Pyrgeometers Kipp&Zonen CGR4 (Infra-red) one mounted on the top of the aircraft (measuring downwelling radiances) and one mounted on the bottom of the aircraft (measuring upwelling radiances) (NEW acquisitions to substitute the former Eppley instruments on ATR-42); -Dew/Frost-point “General Eastern” hygrometer (Dewpoint temperature range -75 -/+50°C); -Rosemount 1201 & Rosemount 1221,to measure static (0-1035 hPa) and dynamic pressure (0-85 hPa); -CIMEL CLIMAT: Radiative temperature 3 wavelengths (-50 – 400°C); -5-port turbulence probe; -Inertial Navigational System + Global Positioning System: Wind component ± 256 kts ± 180 deg;

Other constraints: Flight paths require permission from the Portuguese Institute for civil air traffic control (I.N.A.C). The sampling process might be turbulent, not recommend for researchers highly susceptible to motion sickness. Recent, informal discussions with civil air traffic controllers about the i-WAKE2 proposed flight plans (from NAV-Madeira: alerted for two important issues: (i) Wednesdays afternoon are the calmer air-traffic days; (ii) When in conflict commercial flights will have priority over EUFAR research flights.

Project website:

Scientific contact


PI email:

PI website: