3.1.1 Wind detection

Since ALWIN is targeted at large-scale airports, it is likely that there will already be Doppler radars and/or lidar sensors installed for low-level windshear alerts. Conventional low-level windshear alerts are generated when airport-based Doppler radars and/or lidar detect significant changes in headwinds within alert areas established around the departure and arrival paths (Table 4 and Fig. 6)^{(Reference Matayoshi, Iijima, Yamamoto and Fujita4)}. These alerts are communicated to pilots in real-time by air traffic controllers via radiotelephony so that pilots can take immediate action to avoid the hazard. Alert threshold values of 20kn or 30kn can detect windshear of sufficient severity posing a high risk of an accident even if pilots were to take corrective action on encountering the phenomenon. Thus, existing systems provide only the minimum necessary windshear information: the coarse position of the windshear and the peak value of the headwind change. This information is insufficient to support normal operations and increase operational efficiency by reducing the number of go-arounds.

Table 4

Types and thresholds of low-level windshear alert

Figure 6.

Target areas of low-level windshear alert^{(Reference Matayoshi, Iijima, Yamamoto and Fujita4)}.

Figure 7.

Relationship between windshear intensity and flight control difficulty during final approach^{(Reference Matayoshi, Iijima, Yamamoto and Fujita4)}.

We therefore conducted a flight crew survey^{(Reference Iijima, Matayoshi and Yoshikawa5)}, flight crew interviews and reviews of flight data from which we identified items of wind information that would be relevant to the approach and landing phases of flight and useful for operations. The identified wind detection information includes not only headwinds but also the crosswind component and gust information on the approach and landing flight path, as shown in Table 3. In addition, analyses of go-around events have revealed that aircraft attitude disturbances due to turbulence encounters below 500ft with significant crosswind as well as headwind components are a major cause of such events^{(Reference Iijima, Matayoshi and Yoshikawa5,Reference Matayoshi, Iijima, Yoshikawa and Ushio11)} . Such turbulence that causes aircraft attitude changes is typically topographically induced or building wake and has small spatial structures (on the order of 100m) not covered by existing windshear detection systems that only detect disturbances with larger spatial structures (on the order of 1km).

3.1.2 Windshear detection thresholds

To support normal operations, the ALWIN windshear detection thresholds were reduced compared to existing windshear alert (WSA) and microburst alert (MBA) criteria (Table 3, with detail data in Table 4) in order to alert the presence of windshear that is strong enough to force a go-around or other operational disruption but not necessarily strong enough to be a flight safety hazard. In order to clarify a so-called ‘yellow signal’ (a caution rather than alert) wind situation, in which pilots will be able to make a successful approach and landing with moderate to high workload, the threshold value was determined based on our previous study on the effect of windshear on flight control difficulty during the landing phase^{(Reference Matayoshi, Iijima, Yoshikawa and Ushio11)}. Figure 7 shows the relationship between windshear intensity on the approach path and flight control difficulty (pilot subjective rating) for 154 landings at Shonai Airport, a Japanese regional airport^{(Reference Matayoshi, Iijima, Yamamoto and Fujita4)}. Although there is a large variation in the results, pilots felt that flight control was difficult when the windshear intensity exceeded 10kn/100ft^{(Reference Matayoshi, Iijima, Yamamoto and Fujita4)}. Accordingly, ALWIN indicates the presence of turbulence when the difference between the gust (peak) and average wind speeds exceeds 10kn. This turbulence detection threshold is the same as the gust reporting criteria of airport surface winds. This turbulence alert feature is a novel feature of ALWIN, which can detect turbulence with small-scale spatial structures that could cause aircraft attitude disturbances^{(Reference Matayoshi, Iijima, Yamamoto and Fujita4)}.

3.1.3 Information content and display design

1. (a) ACARS text display

Since ALWIN is targeted at large-scale airports with multiple runways and traffic exceeding 30 airline flights per day, manual ACARS transmission by ground operations staff is considered to be unrealistic. For this reason, ALWIN is assumed to be operated in pull mode rather than push mode as mentioned in Section 2.0.

The design of the ACARS text display is a highly important issue for the ALWIN system because the flight crew must make a rapid but accurate situation assessment of wind disturbances from the ACARS text during the high workload approach flight phase, while the character count of each ACARS text message is limited. We describe design concept of the ALWIN ACARS text message in the following paragraphs.

Table 5

Correlation coefficient analysis results between subjective flight control difficulty altitude range and subjective workload rating^{(Reference Iijima, Matayoshi, Fujita and Yamamoto13)}

The ALWIN ACARS message format provides not only the same critical microburst alert and windshear information as conventional systems⁽¹²⁾ but also information that can support landing planning and decision-making such as Pilot Information Reports (PIREPs). We have proposed a novel ACARS message format, which is the world’s first routine operational ACARS message that contains a spatial graph rendered as text^{(Reference Iijima, Matayoshi and Yoshikawa5)}. We hypothesise that this headwind graph will intuitively provide pilots with better awareness of headwind changes with altitude. Based on the concept, the ALWIN ACARS text message provides two types of advisory information for aircraft operations support: Final (FNL) ACARS and Airport (APT) ACARS, as shown in Fig. 3(a)

• FINAL (FNL) ACARS

The FNL ACARS message format focuses mainly on low-level (altitude) wind information on the final approach path to improve the resolution of the altitude and wind speed information provided to flight crews. The low-level wind information is limited to below 1,000ft, as shown in Table 3, because general flight crews comments (which were obtained from a previous study) indicated that it is necessary to obtain wind information at least below 1,000ft to allow pilots to satisfy the stabilised approach criteria for continuing an approach. Moreover, the FNL ACARS message provides high-resolution wind information below 200ft because subjective ratings of high flight control difficulty from our flight crew survey^{(Reference Iijima, Matayoshi, Fujita and Yamamoto13)} correlated especially to altitudes below 200ft, as shown in Table 5. In addition, the FNL ACARS message provides not only average winds but also wind speed variation, shown as ‘< >’ in Fig. 3(a)(1), so that pilots can recognise rough air and/or turbulent conditions and the potential existence of updrafts and downdrafts.

Based on this concept, profiles of headwind and crosswind components along the final approach path are depicted by graphical text formats, as shown in Fig. 3(a)(1). Heights at which vertical windshear or turbulence exist are automatically detected and indicated in the ‘REMARKS’ area and on profile graphs. Vertical windshear is indicated when the wind speed change of either headwind or crosswind components exceeds 10kn/100ft, as mentioned in Section 3.1.2.

ALWIN’s FNL ACARS is therefore designed to support decision-making when determining the target approach speed, runway selection in the case of multiple parallel runways, and flight control strategy under windshear/turbulent /crosswind conditions.

• AIRPORT(APT) ACARS

The APT ACARS message format shows the vertical profile of averaged winds over the airport in a tabular format that mainly focuses on global wind information up to high altitude. Notable changes in wind speed or direction are automatically detected and indicated in the table. The APT ACARS message format is designed to support flight planning for the timing of configuration changes (landing gear and flaps extension) by providing rough altitude wind direction/speed change information to flight crews.

Ground-based operators can retrieve the ALWIN text information by clicking the ‘ACARS text conversion’ button on the ALWIN web screen shown in Fig. 4(a). ALWIN text information is thus accessible in two ways: either as a display of text information on the ALWIN web screen or as text information downloaded directly from a server over a data communication network.

1. (b) Web display

ALWIN provides two types of advisory information for operations support staff through web pages that can be delivered over the Internet: a FNL screen and an APT screen, as shown in Fig. 4(a). The design concepts of the FNL and APT screens are as same as the ALWIN ACARS text message formats shown in Table 3.

ALWIN screen displays have been provided as a part of the JMA’s MetAir web-based aeronautical weather information service⁽¹⁴⁾ since April 2017. Registered users can browse ALWIN information using ordinary web browsers.

• FNL screen

In accordance with the design concept, the FNL screen shows profiles of headwind and crosswind components along the final approach path in graphical and tabular formats, as shown in Fig. 4(a)(1). Heights at which vertical windshear or turbulence exist are automatically detected and indicated in the Information Area and on profile graphs as coloured bands. Vertical windshear is indicated when the change in either headwind or crosswind components exceeds 10kn/100ft, as mentioned in Section 3.1.2.

• APT screen

As with the APT ACARS message format, the APT screen shows the vertical profile of averaged wind over the airport in a tabular format that mainly focuses on global wind information up to high altitude. Notable changes in wind speed or direction are automatically detected and indicated as ‘SPD CHG’ or ‘DIR CHG’ in the table. The Information Area shows wind shear line, microburst and temporal wind change trends. This wind shear line and microburst information thresholds use the conventional criteria shown in Table 4 developed by the JMA.

The locations of windshear lines and microbursts are also shown in a plan-view ‘map’ format and overlaid onto the horizontal distribution of winds and radial wind speeds. Since this ‘map’ format aims at improving the wind situational awareness of ground-based operators, it is shown only in the web display of APT screen, not in the APT ACARS message.

Both the FNL and APT screens can provide historical wind information for up to 12 hours previously to allow operators to monitor wind changes over time.

3.2.1 Sensor coverage

LOTAS provides information on low-level wind disturbances (windshear, turbulence, etc.) and information to support approach planning. Based on the operational concept outlined in Section 2.2, Doppler radar and lidar sensors are used to detect atmospheric disturbances under all weather conditions, areas of low visibility due to rain and snow, and wind speed and direction, while precipitation intensity is obtained from Doppler radar.

Since the altitudes and distances from the airport at which low-level wind disturbances occur differ for each airport due to factors such as local terrain, we survey the altitudes and distances at which low-level wind disturbances are judged to have a significant impact on flight operations based on wind observation data, flight parameters (e.g. airspeed) of specific aircraft types, and interviews with flight crews and airline operations staff. Regarding the weather environment of LOTAS operations at Shonai Airport, for example, we determined that information should be provided on low-level wind disturbances below 500ft and radar echoes below 2km altitude within a 10NM radius of the airport.

It should be noted that if wind observations are required at higher altitudes or over a wider area, the installation and maintenance costs of LOTAS will increase.

3.2.2 Display design

At present, the design of LOTAS assumes push mode operation to avoid the cost of integrating LOTAS into the airline’s ACARS ground system. This is considered feasible because the operations staff workload will be lower at smaller scale airports with only a few scheduled daily flights. If the operations staff decides to advise a flight crew of the wind situation, they generate the LOTAS ACARS message text by clicking the ‘ACARS text conversion’ button in the wind information web display shown in Fig. 4(b)(1). The generated LOTAS ACARS text, shown in Fig. 3(b), is then manually copied to the company’s ACARS terminal to uplink the message or verbally communicated to the flight crew.

The design of the web display is highly important for the LOTAS system because the operations officers might have to assess wind disturbances and favourable aircraft landing timing in rapidly changing conditions with strong precipitation on the approach route. Moreover, concise and accurate advice must be communicated to flight crews during the high workload approach phase of flight.

Considering these requirements, the LOTAS display^{(Reference Iijima, Matayoshi and Yoshikawa5)} prototype introduces graphical wind information and radar echo displays and a specific ACARS message text derived from a newly developed severity metric (estimated landing difficulty)^{(Reference Matayoshi, Iijima, Yoshikawa and Ushio11)}, headwinds and other information as shown in Figs 3(b) and 4(b).

The wind information display mainly provides alerts of low-level wind disturbances. The radar echo display, which shows precipitation intensity, is intended to be used for decision support to avoid areas of precipitation that will affect the aircraft’s approach and to plan the landing timing when there are areas of precipitation within 10NM of the airport. The proposed ACARS message text, which provides low-level wind conditions to flight crews, is designed to improve wind situational awareness, the ability to predict the future wind situation in dynamic conditions and in-flight approach and landing planning, similarly to ALWIN.

1. (a) Web display

   • Wind Information Display

The wind information display presents the following information: alert messages, a landing path headwind graph profile (HW graph) and a wind direction/speed (headwind component) data table (wind data table), as shown in Fig. 4(b)(1). These are presented side-by-side on the same screen to enable operators to easily assimilate the contents of alerts by the following process. The operator first obtains summarised alert information from the alert message area, grasps the general wind situation from the graphical information presented in the HW graph and finally confirms precise wind data from the numerical information in the wind data table. Based on this process, we hypothesise that ground staff will find it easier and more appropriate to use the system in push mode, rather than the pull mode used with ALWIN.

As mentioned earlier, the display also includes newly created severity information that reflects the impact of wind disturbances on aircraft state and flight control difficulty, and the ‘estimated landing difficulty’ metric^{(Reference Matayoshi, Iijima, Yoshikawa and Ushio11)} for low-level wind disturbances to be presented in the alert messages area, shown as ‘L/D COND’ in Fig. 4(b)(1). The L/D COND for each aircraft type has three levels, ‘RED’, ‘AMBER’ and ‘GREEN’, as defined in Fig. 4(b)(1). Additionally, the alert message area provides landing difficulty trend information (e.g. ‘TREND UNCHANGE’), head/tail windshear component (‘GAIN/LOSS’) and altitude information.

Airline operations staff sometimes forecast low-level wind disturbances based on the previous few hours of observations. Therefore, the LOTAS display has a function to provide past information (called ‘Warning History’ in this paper) of L/D COND, wind direction and speed information over the previous several hours, as shown in Fig. 4(b)(3). The information is accessed by the Warning History button shown in Fig. 4(b)(1)

• Radar Echo (Precipitation, Snow intensity) Display

As mentioned above, preparation for landing requires not only information on low-level wind disturbances but may also require radar echo information on the intensity of rain or snow. Conventional radar echo information, for example⁽¹⁵⁾, is difficult to use for landing-timing decisions – its poor spatial resolution and low update rate (e.g. 10 minutes) make it hard for flight crews to decide when to commence an approach in a rapidly changing weather situation.

We therefore present a new radar echo information display with higher spatial resolution and update rate (e.g. 2 minutes) focusing on low altitudes (below 2km) in the approach area. To further support landing-timing decision-making, we propose a radar image that shows both the ‘current observation’ and a ‘10-minute (10-min) prediction’ of radar echoes superimposed on an approach chart, as shown in Fig. 4(b)(2), since it takes approximately 10 minutes from an aircraft leaving the ‘high station’ (initial approach fix) of the airport to starting its final approach and landing. By comparing the two images, an operations officer will be able to estimate what the situation will be 10 minutes later, that is during the landing phase, before the aircraft starts its approach, thus enabling better judgment of the landing timing.

1. (b) ACARS text display

The LOTAS ACARS, shown in Fig. 3(b), differs from the ALWIN ACARS display in that it provides an estimated landing difficulty metric^{(Reference Matayoshi, Iijima, Yoshikawa and Ushio11)} (L/D COND). Although the LOTAS ACARS does not include the crosswind graph and wind variation notation ‘< >’ of the ALWIN ACARS, it is possible to include such information by customising the message according to the specific local weather environment of the airport. At Shonai Airport, used in the operational evaluation of LOTAS described in Section 4.0, the LOTAS ACARS message omitted the crosswind graph because the headwind shear was considered to be most problematic local weather phenomenon and was therefore prioritised.

3.3.1 Wind detection

As shown in Table 1, although SOLWIN’s observation area is limited, it has the advantage of providing vertical wind speed information, that is updrafts and downdrafts, which the other systems do not. In order to support aircraft operations effectively, vertical wind speeds are presented in units of ft/min (fpm) rather than kn so that flight crews can more easily estimate the effect on the aircraft’s rate of climb or descent.

While horizontal wind information (wind direction/wind speed) reported to pilots is given as two-minute average values, an evaluation of SOLWIN at Oita Airport found that this was too long for vertical wind fluctuations, so the one-minute averages are given to allow severe updrafts and downdrafts to be indicated.

We hypothesise that displaying vertical speed change when significant updrafts and downdrafts exist will be helpful to pilots not only in the case of steady vertical winds but also for dealing with short-scale phenomena such as building wake, small mountain wave and hanger wave. (‘Hanger wave’ is turbulence in the wake of a hanger that is located near a runway in strong wind and can disturb aircraft just prior to touchdown.)

3.3.2 Windshear detection thresholds

SOLWIN uses the same headwind and crosswind threshold values to indicate the presence of windshear as ALWIN and LOTAS, namely 10kn/100ft. The values used in the operational evaluation presented in this paper are shown in Table 6 and were based on interviews with flight crews. These interviews were important because expressing vertical wind speed in fpm is novel. During the present SOLWIN evaluation phase, the presence of updrafts or downdrafts is displayed when the constant estimated vertical wind speed exceeds 300fpm. These values were selected considering a nominal rate of descent of approximately 700fpm for a Boeing 737-sized aircraft on final approach and the operational requirement to establish a stabilised approach with a descent rate of less than 1,000fpm; that is, a downdraft of more than 300fpm will cause the stabilised approach descent rate criterion of 1,000fpm to be exceeded.

Based on this design concept, the windshear detection threshold values and priority of indication shown in Tables 6 and 7 were chosen.

3.3.3 Information content and display design

1. (a) ACARS text display

At the beginning of the system feasibility study for SOLWIN, although we assumed that SOLWIN would be used mainly in push mode, an airline expressed a desire for pull mode operation as its expectation of the effectiveness of SOLWIN increased.

Table 6

Windshear detection thresholds in SOLWIN

HW: headwind, XW: crosswind, VW: vertical wind

Table 7

SOLWIN information provision priority

DN DFT: downdraft, UP DFT: updraft, HW: headwind, XW: crosswind, VW: vertical wind

The design of the SOLWIN ACARS text display therefore also became an important issue. In particular, it was required that the new display design support intuitive awareness of up draft and down draft information. However, the size of the ACARS text display area is limited by the airborne ACARS software specification of 23 lines vertical by 39 letters horizontal, and flight crews also requested to see headwind and crosswind graphs. Therefore, simple indications of ‘UP’ and/or ‘DN’ were shown beside the headwind graph to allow flight crews to recognise intuitively the altitude areas in which updrafts or downdrafts exist, as shown in Fig. 3(c). As shown in the figure, the headwind graph only shows heights up to 300ft above the runway threshold due to the limitations of the Doppler SODAR sensor.

1. (b) Web display

The SOLWIN web display, shown in Fig. 4(c), is based on the same concept as the ACARS text display design described above.

Information is provided to ground-based operations staff in a tabular format to allow them to easily recognise wind situation tendencies, and because we suppose that SOLWIN will be used in both pull and push modes of operation.

6.3.1 ALWIN

This result, mentioned in Section 5.3, indicates that flight crews regard high altitude information as important, and we hypothesise that APT ACARS supports flight planning (e.g. timing of configuration changes). In fact, one flight crew at Haneda Airport commented directly that they determined the target approach speed and timing of configuration changes from APT ACARS information. It is considered from a number of flight crew comments that the reason why the ALWIN system was rated as more useful at Narita Airport than at Haneda, as mentioned in Section 5.1, was due to the missing high-altitude information from APT ACARS messages.

Considering the aspects of ‘not getting information just before L/D (landing)’ and ‘not responding real time (changing wind) situation’, since flight crews could not obtain ALWIN ACARS just before landing due to the ‘sterile cockpit’ concept^{(Reference Sumwalt16)}, it is desirable to give future consideration to the real-time provision of ALWIN-type information.

6.3.2 SOLWIN

Since SOLWIN cannot measure wind data on the approach flight path, unlike ALWIN, a difference between the actual encountered wind situation and SOLWIN wind information is considered unavoidable.

Nevertheless, there were almost unanimous flight crew requests for SOLWIN implementation, as shown in Fig. 12, and we obtained positive comments shown in Table 11 as mentioned in Section 5.2. These results indicate that SOLWIN will be useful for regional airports, although some issues remain to be addressed in the future.

From the result of ‘difficult to understand info. (information)’, it is considered that the presentation of updraft and downdraft information is novel and crews had no experience as to how it could be applied, and that the ACARS message text itself may also be hard for flight crews to understand. A method for presenting the new severity metric message and how to apply it are future issues.

For the results of ‘others’, we discuss SOLWIN’s observation range limitation in Section 5.6.

6.5.1 LOTAS

1. (a) ACARS text display for flight crews

These results mentioned in Section 5.5.1 support our hypothesis regarding the ACARS text display design concept that converting spatial information into a textual graph promotes intuitive understanding of headwind changes with altitude. We consider that an intuitive understanding is highly important, as mentioned in the design concept, because a rapid but accurate situation assessment of wind disturbances is required during the high workload approach flight phase.

‘L/D COND’ (estimated flight control difficulty) was rated as effective information by fewer flight crews than the ‘Headwind Graph’. While the L/D COND indication is accurate (that is, it agrees with actual conditions) nearly 70% of the time^{(Reference Matayoshi, Iijima, Yoshikawa and Ushio11)}, it is possible that fewer crews regarded it as useful for aircraft operations because the information itself is novel and they had no experience as to how it could be applied. Moreover, it is possible that the meaning of L/D COND was not always sufficiently understood, since some crews asked about it after landing. Because of this, a method for presenting the new severity metric message and understanding how to apply it are future issues to be addressed.

1. (b) Web display for operation officers

These results mentioned in Section 5.5.1 and in Fig. 18 indicate the importance of predictive and historical information for landing-timing decision support. Additionally, it is considered that the flight control difficulty information (L/D COND) provided direct hints for ground-based operations staff to advise flight crews of the wind situation, as we hypothesised in the LOTAS operational concept in Section 2.2. Furthermore, one operation officer commented that information on changes of L/D COND provided by the ‘Warning History’ shown in Fig. 4(b) might be applicable to deciding whether or not to start the approach because past information on L/D COND may help the operator to predict its future state.

Thus, it is considered that LOTAS display and functions are effective for low-level wind disturbance alerting and landing-timing decision support. The novel L/D COND information was considered by operation officers to be effective for providing a direct cue with which to advise flight crews on the wind situation, although use of the same information by flight crews is a topic for future discussion.

6.5.2 SOLWIN

From the results in Fig. 19 and Table 10, it is considered that the proposed SOLWIN ACARS message is almost as effective for supporting aircraft operations as the other displays. However, Fig. 19 suggests that flight crews were not able to use updraft and downdraft information. On the other hand, the evaluation of ‘Wind DIR/SPD (Wind Direction and Speed)’ as the most useful information item corresponds to the evaluation result of LOTAS mentioned above and ALWIN in our previous study results. We suppose that since flight crews are familiar with the notation of wind direction and speed, which are normally provided by air traffic control and in other weather information, they were easily able to use this information.

Since flight crews mainly control not vertical speed but airspeed in windshear/turbulence conditions, presenting updraft and downdraft information as airspeed loss or gain information (in kn) rather than vertical speed might make airspeed control easier. Moreover, if updraft and downdraft information were to be included with such airspeed gain/loss information, the ACARS display would be further simplified and could provide more strategic information because flight crews would be able to use the SOLWIN information directly for airspeed control. Also, if the ACARS display were further simplified, the ‘difficult to understand info. (information)’ criticism of SOLWIN mentioned in Section 5.3 might be improved. How updraft and downdraft information should be displayed is therefore a subject for future discussion.

6.6.1 LOTAS

As mentioned in Section 5.5.1, it is considered that when predicting the future weather situation from 10 to 20 minutes of historical data, information on changes in ‘L/D COND’ provided by the ‘Warning History’ shown in Fig. 4(b) might be applicable to deciding whether or not to start an approach.

It may therefore be possible to gain further effectiveness by increasing the radar echo prediction time to 20 minutes, enabling better advice from operations officers to crews as to when to start their approaches. This will be possible if the radar observation range is increased.

Summarising the usability of the limited observation area of the LOTAS system mentioned in Section 5.6, although the low-cost Doppler lidar and radar sensors^{(Reference Matayoshi, Iijima, Yoshikawa and Ushio11)} used in this evaluation have only a tenth of the observation range of systems installed at major airports, they have been proven to be effective in providing operationally useful wind information up to 500ft altitude and radar echo information below 2km in altitude within a 10NM radius of the airport. The low-cost of these sensors and their effectiveness demonstrated by this evaluation might accelerate the installation of LOTAS at regional airports that are prone to severe low-level wind disturbances.

6.6.2 SOLWIN

Although SOLWIN’s Doppler SODAR wind observation sensor was installed at the runway threshold in this evaluation, it is necessary to install the sensors at locations that are highly correlated with turbulence or windshear on the landing path, for example due to local topography. If sensors can be installed at such locations where their data will be most effective, wind situational awareness might be improved even if only wind data below 300ft are available. It is therefore recommended that unless the Doppler SODAR observation range can be improved, the SOLWIN sensor locations should be determined at each airport by investigating the correlation between SOLWIN wind measurement data and flight data on the landing path.

There were also comments requesting installation of Doppler SODAR not only for the principal runway direction but along the landing paths at both ends of the runway. In fact, during evaluations, even if a flight crew landed not on runway 01 but on the reciprocal runway 19, where SOLWIN information was not provided, they still used the SOLWIN system. Questionnaire data showed that approximately 20% of flights that used SOLWIN landed on runway 19, and it is considered that in such cases, flight crew estimated the runway 19 approach wind situation from the SOLWIN information for runway 01. While it will be possible to provide wind information for both ends of a runway by installing sensors under both approach paths, this will increase cost and might negate SOLWIN’s low cost advantage shown in Table 1. We suppose that cost benefit analyses of not only the SOLWIN system but also LOTAS and ALWIN are future issues.