Narrative:

Winds at roa were forecast to be above 20 knots at the surface with gust factors above 18 knots. The weather otherwise was severe clear with clear skies and visibilities well above 10 SM. We were prepared for and briefed the flight attendant and the passengers for a turbulent approach and landing. Upon arrival at roa the winds were still strong with LLWS warnings on the ATIS but aligned with runway 34. It was very turbulent in the last 3000 ft with significant airspeed excursions in the gusts. On about a 1.5 mile final the tower issued a windshear alert for a 15-knot loss on a one mile final. Almost simultaneously we experienced a 15-knot gain in airspeed; with; shortly thereafter; an increasing performance windshear warning. We rejected the landing and went around. We attempted another approach to the same runway but this time received a 20-knot loss warning from tower at about the same location with the same result - an increase in airspeed followed by an increasing performance windshear warning - we went around. On this go-around we evaluated the fuel required to divert to ZZZ and determined that we would divert to ZZZ if we could not get in. We asked the tower about landing in a crosswind on runway 24 - specifically if other aircraft had used that runway and if the tower had received any windshear alerts from that area. Receiving an affirmative of the former with no windshear reports and a negative about windshear alerts from the ground-based sensing system; we attempted a third approach to runway 24. Unfortunately this ride was rougher than the previous ones and this time we received a decreasing performance windshear alert from the onboard system. We went around a third time. Doing a quick assessment on fuel; the trending of the conditions; the effects on passenger comfort of the three previous attempts; and our own nerves; we decided to divert to ZZZ. Once on the ground we used our windshear go-around guidance of +/- 15 knots of instantaneous airspeed change to determine that we would stay in ZZZ until the reported gust factors were down to 10 knots in order to give us a good chance to get in. We monitored the conditions by calling the roa ASOS phone number every 15 minutes. We waited about two hours for conditions to subside; then loaded up and made it in to roa even though the ride was still very turbulent. Shortly after we opened the door; the tower issued a windshear alert for a 20-knot loss on the approach end of the runway we had just landed on. On the subsequent departure; climbing through about 4000 ft MSL we flew into a mountain wave and experienced a climb rate of up to 5800 fpm. Windshear and very turbulent conditions created an environment where the airspeed deviations were so frequent and random that cognitively we were saturated during the approach. A contributing factor was an approach controller that repeatedly wanted to immediately know our intentions even though we were in the middle of our go-arounds. On the ground at ZZZ; I gathered the crew for a debrief which is my custom after a significant event like this. We identified a communication breakdown from the communication of trends that I identified (at a much lower workload than my flying first officer) that my first officer could not perceive because of his task saturation trying to fly the aircraft in the turbulent go-around. (He didn't see a large rapid airspeed increasing trend vector on a couple occasions) the insistent controller created time pressure to complete the circuit and attempt another approach. After the second go-around; I told him we needed to fly a box pattern and evaluate our options to get him off our backs for a moment. We decided at that time if the third attempt was unsuccessful that we would divert to ZZZ. Having made that decision ahead of time reduced the subsequent divert workload somewhat. On the divert my first officer initially accelerated to normal cruise speed; but I told him we needed to slow down to create some time to catch up. I don't think he was fully aware of how close ZZZ was because we were still in turbulence and he was still loaded up. When we arrived in the area; the approach controller asked us if we had the airport in sight. We did; but I told my first officer I wanted to stay above 10;000 ft until we had our landing numbers; briefed the approach; and had taken a minute to ensure we were caught up with events. I instructed him to ask for a box pattern which he obtained. We had the fuel so we took our time; accomplished all tasks; dotted the I's and crossed the T's; then took a minute to take a breath and look around to ensure that everything needed to conduct the approach was done; including briefing the flight attendant; passengers; and dispatch. Once we agreed that everything was done and we were mentally caught up; I okayed the descent and approach. The rest of the divert proceeded without further incident.the main reason I submitted this as soon as possible was to possibly spur a conversation about dispatch decisions. I know we often depart with the intent to 'take a look' when conditions are marginal - after all; conditions often change significantly while in flight. On this day the high winds were driven by a strong low pressure center (altimeter setting of 29.42 in fwa the previous day) located in the eastern ohio valley and were widespread over the appalachians and forecast to remain so during the day. It occurred to me while we were on the ground at ZZZ that the forecast gust factors were well above our windshear go around criteria of an instantaneous airspeed change of 15 knots and plus/minus 500 fpm. In a situation like this; in mountainous terrain where we know the conditions significantly exceed our windshear go-around guidance and are not likely to change; should this modify our usual approach to 'take a look' or should we treat conditions like this the same as an extensive thunderstorm complex hovering over the destination. Ergo - should our policy be that we wait for improved conditions before launching? Should we use forecast and reported gust factors of above 15 knots as a no-go decision criteria for launching the flight?

Google
 

Original NASA ASRS Text

Title: During multiple visual approaches to ROA; flight crew received windshear alerts from both ground-based sensors via ATC and on-board detection equipment. The crew diverted after receiving the same result on the third approach. The Captain suggested that ATC delay questioning crews about their intentions immediately after a go-around; especially in turbulent/windshear conditions.

Narrative: Winds at ROA were forecast to be above 20 knots at the surface with gust factors above 18 knots. The weather otherwise was severe clear with clear skies and visibilities well above 10 SM. We were prepared for and briefed the FA and the passengers for a turbulent approach and landing. Upon arrival at ROA the winds were still strong with LLWS warnings on the ATIS but aligned with runway 34. It was VERY turbulent in the last 3000 ft with significant airspeed excursions in the gusts. On about a 1.5 mile final the tower issued a windshear alert for a 15-knot loss on a one mile final. Almost simultaneously we experienced a 15-knot gain in airspeed; with; shortly thereafter; an increasing performance windshear warning. We rejected the landing and went around. We attempted another approach to the same runway but this time received a 20-knot loss warning from tower at about the same location with the same result - an increase in airspeed followed by an increasing performance windshear warning - we went around. On this go-around we evaluated the fuel required to divert to ZZZ and determined that we would divert to ZZZ if we could not get in. We asked the tower about landing in a crosswind on Runway 24 - specifically if other aircraft had used that runway and if the tower had received any windshear alerts from that area. Receiving an affirmative of the former with no windshear reports and a negative about windshear alerts from the ground-based sensing system; we attempted a third approach to Runway 24. Unfortunately this ride was rougher than the previous ones and this time we received a decreasing performance windshear alert from the onboard system. We went around a third time. Doing a quick assessment on fuel; the trending of the conditions; the effects on passenger comfort of the three previous attempts; and our own nerves; we decided to divert to ZZZ. Once on the ground we used our windshear go-around guidance of +/- 15 knots of instantaneous airspeed change to determine that we would stay in ZZZ until the reported gust factors were down to 10 knots in order to give us a good chance to get in. We monitored the conditions by calling the ROA ASOS phone number every 15 minutes. We waited about two hours for conditions to subside; then loaded up and made it in to ROA even though the ride was still very turbulent. Shortly after we opened the door; the tower issued a windshear alert for a 20-knot loss on the approach end of the runway we had just landed on. On the subsequent departure; climbing through about 4000 ft MSL we flew into a mountain wave and experienced a climb rate of up to 5800 fpm. Windshear and very turbulent conditions created an environment where the airspeed deviations were so frequent and random that cognitively we were saturated during the approach. A contributing factor was an approach controller that repeatedly wanted to immediately know our intentions even though we were in the middle of our go-arounds. On the ground at ZZZ; I gathered the crew for a debrief which is my custom after a significant event like this. We identified a communication breakdown from the communication of trends that I identified (at a much lower workload than my flying FO) that my FO could not perceive because of his task saturation trying to fly the aircraft in the turbulent go-around. (He didn't see a large rapid airspeed increasing trend vector on a couple occasions) The insistent controller created time pressure to complete the circuit and attempt another approach. After the second go-around; I told him we needed to fly a box pattern and evaluate our options to get him off our backs for a moment. We decided at that time if the third attempt was unsuccessful that we would divert to ZZZ. Having made that decision ahead of time reduced the subsequent divert workload somewhat. On the divert my FO initially accelerated to normal cruise speed; but I told him we needed to slow down to create some time to catch up. I don't think he was fully aware of how close ZZZ was because we were still in turbulence and he was still loaded up. When we arrived in the area; the approach controller asked us if we had the airport in sight. We did; but I told my FO I wanted to stay above 10;000 ft until we had our landing numbers; briefed the approach; and had taken a minute to ensure we were caught up with events. I instructed him to ask for a box pattern which he obtained. We had the fuel so we took our time; accomplished all tasks; dotted the I's and crossed the T's; then took a minute to take a breath and look around to ensure that everything needed to conduct the approach was done; including briefing the flight attendant; passengers; and dispatch. Once we agreed that everything was done and we were mentally caught up; I okayed the descent and approach. The rest of the divert proceeded without further incident.The main reason I submitted this ASAP was to possibly spur a conversation about dispatch decisions. I know we often depart with the intent to 'take a look' when conditions are marginal - after all; conditions often change significantly while in flight. On this day the high winds were driven by a strong low pressure center (altimeter setting of 29.42 in FWA the previous day) located in the eastern Ohio valley and were widespread over the Appalachians and forecast to remain so during the day. It occurred to me while we were on the ground at ZZZ that the forecast gust factors were well above our windshear go around criteria of an instantaneous airspeed change of 15 knots and plus/minus 500 fpm. In a situation like this; in mountainous terrain where we KNOW the conditions significantly exceed our windshear go-around guidance and are not likely to change; should this modify our usual approach to 'take a look' or should we treat conditions like this the same as an extensive thunderstorm complex hovering over the destination. Ergo - should our policy be that we wait for improved conditions before launching? Should we use forecast and reported gust factors of above 15 knots as a no-go decision criteria for launching the flight?

Data retrieved from NASA's ASRS site and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.