Gust Factors in Aerodrome Weather and Climate Assessment

Abstract

Wind gustiness at airports, which is generally determined using gust factors, is impactful across a range of considerations from piloting to airport planning. Yet advisory materials to help assess their quality and representativeness, particularly for aviators, are limited. To address this, a climatological analysis of both gust factors is conducted using Automated Surface Observing System (ASOS) wind observations. Data for multi-year periods at selected airports in the United States are used to assess their site representativeness and for turbulence attribution purposes. Both gust factors vary by direction in response to local terrain features and nearby obstructions and are generally not well correlated with each other. The meteorological gust factor is shown to be more responsive to local obstructions in proximity to the ASOS systems. Excluding lower gusts leads to a marked improvement in the correlation between the two gust factors. Due to ASOS’s siting limitations, attributing observed gustiness to turbulence from nearby terrain or structures is difficult. The gustiness is often localized and may not represent conditions across the full airport. Excluding lower gusts increases the aviation gust factor’s sensitivity to local obstructions. This suggests that obstructions may play a meaningful role in shaping the higher observed gust factors. The potential exists to provide pilots and other users of this data with site- and direction-specific metadata regarding observed gustiness, thereby improving situational awareness.

1. Introduction

Wind gustiness is important in piloting for many fixed-wing and rotorcraft operations. Aircraft accidents in mountainous or high elevation locations with impactful gusts include mountain wave-related and on-airport crosswind cases [1]. The aviation gust factor, also referred to as the gust spread, is defined as the difference between the sustained wind speed and the gust [2]. An often-recommended technique for fixed-wing aircraft pilots during the short final approach is to add one-half the gust factor to the normal approach speed [2,3,4]. Little, if any, guidance is available to aviators regarding the assessment of the quality or representativeness of observed gust factors.

The subset of gusty conditions that may impact aviators at low altitude can include those created by mechanical turbulence caused by local terrain, land-cover, and obstructions such as buildings [5]. Compared to gusts, documented aircraft incidents that specifically mention or infer mechanical turbulence in the vicinity of airports are relatively infrequent. National Transportation Safety Board (NTSB) reports, which note mechanical turbulence from terrain or buildings in fixed-wing aircraft incidents on takeoff or landing, are included in

Table 1. NTSB reports with noted mechanical turbulence near airports for fixed-wing aircraft.

NTSB Case Number	Location	Date	Synopsis
ERA23LA187	Cedar Key, Florida	7 April 2023	Mechanical turbulence encountered crossing the runway 23 threshold.
ANC19CA002	Gambell, Alaska	23 October 2018	Aircraft was flown higher than the glideslope in anticipation of mechanical turbulence expected by the pilot when winds are out of the east.
CEN18LA172	Borger, Texas	11 May 2018	Pilot reported strong updrafts in association with unusual landforms on short final.
CHI08CA013	Middlefield, Ohio	4 October 2007	Pilot reported a sudden loss of lift on approach due to mechanical turbulence from trees.
FTW98LA247	Claremore, Oklahoma	28 May 1998	Pilot reported being unable to maintain altitude due to mechanical turbulence from trees off the end of the runway.
FTW85LA241	Fort Worth, Texas	31 May 1985	Mechanical turbulence from nearby buildings on takeoff.

. An incident at the Hutchinson County Airport (BGD), Borger, Texas, for example, notably included reference to “strong updrafts from unusual landforms” with respect to a short final approach on runway 21 [6]. The NASA Aviation Safety Reporting System (ASRS) [7] additionally has reports that include references to mechanical and orographic turbulence in the vicinity of airports. Reports for fixed-wing aircraft are included in

Table A1. NASA ASRS reports related to mechanical turbulence.

ASRS Report Number	Location	Date	Relevant Report Text
1960621	Unknown	12/2022	“This airport is known for rogue gusts as it sits on a hill with the south side of the runway having rising terrain, which seems to create gusts at the numbers on [the] Runway XY approach.” “Much of the time AWOS reports variable wind instead of a direction because it is varying anywhere from 20–40 degrees.”
1911945	Unknown	06/2022	“There are buildings (and the helicopter pad) to the immediate north-east of the runway. Given wind direction, it is likely that the first part of the runway had stable winds, but the latter part of the runway had more abnormal gusts and eddies caused by wind at the time flowing over those structures.”
1910295	Unknown	06/2022	“The gusty conditions interacting with the large buildings of the city led to downdrafts and mechanical turbulence that kept sucking us down as we tried to climb.”
1887173	Unknown	03/2022	“I think it’s possible that the wind that we experienced was higher than what was reported; or possibly the direction of the wind, coming through the hangars and other airport buildings, might have introduced wind shear at the worst possible point in our landing.”
1864945	Unknown	12/2021	“The east side of the airport is where the control tower, some hangars, and trees are. With the winds being out of the east it makes me wonder if we had a weird gust of wind mixed with some mechanical turbulence.”
1760076	Unknown	08/2020	“I wonder if there could be a connection between the CRJ holding short, perpendicular to the numbers and the easterly, variable winds--possibly generating a unique mechanical turbulence which caused the rapid decay in descent so close to the runway”
1509079	Unknown	12/2017	“Rather than return to the airport of departure, which was beyond some hills creating orographic turbulence that could have added to the load factor, I elected to return to a nearby airport that had long, wind-aligned runways”
1420057	Unknown	01/2017	“Trees, hills, and the open area probably produced gusts and directional changes in the wind.”
1383903	Unknown	08/2016	“I believe the combination of the small amount of ice, aileron deflection and mechanical turbulence from buildings on the upwind side of the runway caused the left wing to stall at a higher than normal airspeed, resulting in the uncommanded left roll.”
1363436	Alturas, California (AAT)	06/2016	“In the flare, the aircraft dropped suddenly due to mechanical turbulence created from the wind flowing over the hangars/FBO/trees on the NW side of the airport.”
1340662	Unknown	02/2016	“During the landing, I encountered a gusting wind and mechanical turbulence from the building and terrain on the west side of runway 14”
1071503	Unknown	03/2013	“The crosswind component was 25 KTS and the tailwind component 9 KTS. We chose Runway XX to avoid mechanical turbulence associated with terrain south of the airport.”
1050984	Alaska	11/2012	“At this airport moderate or greater mechanical turbulence can be expected below 1500–2000 feet when there are strong winds out of the northwest or northeast quadrants”
1045076	Melbourne, FL (MLB)	10/2012	“There are buildings (and the helicopter pad) to the immediate north-east of the runway. Given wind direction, it is likely that the first part of the runway had stable winds, but the latter part of the runway had more abnormal gusts and eddies caused by wind at the time flowing over those structures.”
1016862	Unknown	06/2012	“I thought this must have been due to mechanical turbulence associated with the gusty winds (approximately 150 degrees) coming over the terminal.”
989017	Unknown	01/2012	“The wind was gusty and there was some mechanical turbulence from a large hangar across the runway.”
944692	Unknown	04/2011	“The reported winds, later verified with airport operations, were 180 degrees at four knots. However, the winds were coming from the side of the airport where buildings are relatively close to the runway.”
919797	Unknown	11/2010	“We were high as there was orographic turbulence and we were asked by Tower to turn an early base because there was inbound traffic on long final.”
713739	LaGuardia, NY (LGA, East River Exclusion Zone)	10/2006	“4) THE HT AND PLACEMENT OF THE BUILDINGS AND LCL GEOGRAPHY CREATE UNPREDICTABLE LCL WIND AND WX CONDITIONS.”
497557	Buffalo, NY (BUF)	01/2001	“RPTR CLAIMS WINDSHEAR EQUIP AT BUF DOES NOT GIVE ACCURATE READINGS BECAUSE IT IS AMONGST SURROUNDING VEGETATION AND BUILDINGS.”
485620	Mammoth Lakes, California (MMH)	09/2000	“THE RWY IS SURROUNDED BY EXTREMELY RUGGED AND HIGH TERRAIN AND HAS ASSOCIATED TURB AND GUSTS.”
137899	Marquette, MI (MQT)	02/1990	“THE LCL TERRAIN INCLUDES A LARGE ROCK HILL JUST TO THE N OF RWY 08–26 IN MARQUETTE. THIS HILL CREATES MECHANICAL TURBULENCE AROUND IT WHICH MAKES THE WIND VARIABLE ACROSS THE ARPT PROPERTY, AND CREATING SITUATIONS SUCH AS THE “WHITE OUT” WE ENCOUNTERED.”

, though such references do not necessarily indicate that the mechanical turbulence was a contributing factor to the reported event.

The meteorological gust factor, which is often used to characterize surface roughness, is defined as the ratio of the maximum gust to the mean wind speed over a given time interval. There are many applications of this parameter, ranging from tropical cyclone winds to fire weather forecasting [8,9,10,11]. Natural and human-made obstructions can significantly impact the representativeness of observed sustained winds and gusts. Consequently, gustiness exhibits a strong dependence on wind direction, and its spatial footprint may be highly localized. When local obstructions affect wind measurements, the World Meteorological Organization (WMO) and the National Oceanic and Atmospheric Administration (NOAA) recommend siting the wind sensor higher than the standard level of approximately 10 m [12,13]. Gust factors in this context have been used to filter out unrepresentative observations [14]. Harris & Kahl [10] additionally note that the hourly information provided by Automated Surface Observing System (ASOS) for METAR reports is inadequate for site gust characterization and stressed the importance of using the higher resolution 1-minute data.

The objectives of this study are to address the differences between the aviation and meteorological gust factors in terms of their representativeness and the ability to utilize these gust factors for the attribution of turbulence. Representativeness of the gust factors has practical implications for piloting in the context of real-time situational awareness as well as site characterization in climatological assessments. Attribution could provide some background knowledge for aviators, but it may be considered more relevant to addressing causal factors in gust-related issues in the airport vicinity.

2. Data and Methods

2.1. Datasets

The gustiness analysis was performed using six years (2019–2024) of one-minute ASOS observations of the average and corresponding peak wind speed from 20 stations, arbitrarily selected from across the contiguous U.S. (

Table 2. ASOS locations and the percentage of land cover (low, medium, and high intensity) and forest within 3 km of the ASOS location. The land use information was obtained from the National Land Classification Database.

ICAO Airport Code	Location	NLCD % Developed L-M-H, Forest
KABQ	Albuquerque, NM	58.3, 00.0
KAOO	Altoona, PA	14.5, 07.1
KAZO	Kalamazoo, MI	46.5, 05.4
KBGD	Borger, TX	28.9, 00.0
KBUF	Buffalo, NY	84.2, 00.2
KCKV	Clarksville, TN	42.1, 12.7
KDFW	Dallas-Fort Worth, TX	86.9, 01.2
KGEZ	Shelbyville, IN	16.9, 02.4
KGLS	Galveston, TX	40.6, 00.0
KHGR	Hagerstown, MD	41.0, 02.4
KHSV	Huntsville, AL	28.6, 00.5
KIDA	Idaho Falls, ID	42.5, 00.0
KLXV	Leadville, CO	12.2, 54.3
KMHT	Manchester, NH	48.0, 24.6
KONT	Ontario, CA	92.1, 00.0
KOVE	Oroville, CA	09.8, 00.0
KPAE	Everett, WA	64.9, 11.6
KPKD	Park Rapids, MN	19.5, 11.8
KSLC	Salt Lake City, UT	86.1, 00.0
KTCC	Tucumcari, NM	02.3, 00.0

). The 2023 National Land Cover Database (NCLD) data at 30 m resolution were obtained via the Multi-Resolution Land Characteristics (MRLC) Consortium (https://www.mrlc.gov/data (accessed on 14 May 2025)). The NLCD data were used to calculate the combined percentage of land cover that includes low-, medium-, and high-intensity developed regions, as well as the percentage of forested areas within a 3 km radius of the ASOS location. These values, included in Table 2, are meant to give a qualitative assessment of the roughness in the vicinity of the ASOS. The one-minute ASOS data were accessed from the National Centers for Environmental Information (NCEI) https://www.ncei.noaa.gov/pub/data/asos-onemin/ (accessed on 14 May 2025).

2.2. Aviation and Meteorological Gust Factors

Aviation gust factors (GF_avn) and meteorological gust factors (GF_met) were determined from the sustained and peak winds. A traditional approach (Equation (1)) was used, in which the GF_avn is defined as the difference between the peak (U_max) and average wind (U), i.e.,

$${\mathrm{G}\mathrm{F}}_{\mathrm{a}\mathrm{v}\mathrm{n}}=\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{k} \mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d}-\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e} \mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d}={\mathrm{U}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-\mathbf{U}.$$

(1)

U_max is typically based on a window of a few seconds, while the average wind speed is calculated using a 10 min window. Here, the ASOS peak winds are based on a 3 s peak wind for the gust and 2 min window for the average wind speed [15].

The GF_met is determined from the approach of Yu and Chowdhurdy [16], where a peak factor (g), wind speed standard deviation (SD_u), and average wind (U) are used to obtain the difference between the peak and average wind:

$${{\mathrm{U}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-\mathbf{U}=\mathrm{G}\mathrm{F}}_{\mathrm{a}\mathrm{v}\mathrm{n}}=\mathrm{g}·\mathbf{U}{·\mathrm{S}\mathrm{D}}_{\mathrm{u}}.$$

(2)

Dividing Equation (2) by the average wind speed and defining GF_met as the ratio of the peak-to-average wind speed minus one, we obtain the following:

$${\mathrm{G}\mathrm{F}}_{\mathrm{m}\mathrm{e}\mathrm{t}}=\left[{\displaystyle \frac{\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{k} \mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d}}{\mathbf{U}}}-1\right]={\displaystyle \frac{{\mathrm{G}\mathrm{F}}_{\mathrm{a}\mathrm{v}\mathrm{n}}}{\mathbf{U}}}=\mathrm{g}·{\mathrm{S}\mathrm{D}}_{\mathrm{u}}.$$

(3)

Using these definitions, GF_met is simply GF_avn normalized by the average wind speed. In this case, the gust factors approach zero as the peak and average wind speeds converge, whereas in the traditional formulation, the lower limit of the meteorological gust factor is 1. The removal of the offset renders the comparison of the two factors more straightforward. The gust factors were averaged by direction in 10-degree increments when the number of observations was greater than or equal to 10. To test the impact of filtering lower gusts, factors were calculated using different minimum threshold values (1, 5, 10, 15, 20, and 25 kt).

Additionally, gust reporting from ASOS systems differs in terms of what is transmitted via Aviation Routine Weather Reports in METAR and SPECI formats versus the higher temporal resolution one-minute ASOS observations. The METAR/SPECI reports gust reporting has, among other criteria, a minimum allowable gust report of 14 knots [17], whereas the one-minute ASOS wind gust reports the mean wind and gust without restriction at 1 kt resolution. Additionally, the one-minute ASOS reports the wind directions at 1 deg resolution versus the 10 deg resolution in ME-TAR/SPECI reports.

Four of the 20 ASOS locations are used to demonstrate the differences between the average GF_met and GF_avn (Figure 1). Three of the stations (KDFW, KGLS, and KPAE) have discernible impacts from nearby obstacles, while the fourth, KABQ, experiences canyon flows. The analysis is then extended to the remaining 16 stations (Appendix B, Appendix C and Appendix D).

3. Results

The analysis of the aviation and meteorological gust factors first focuses on their directional variability with filtering based on gusts. Motivated by the filtering of gusts in METAR reports, the gust factors are analyzed with varied levels of filtering based on the gust values. The differences between the unfiltered and filtered versions of the average gust factors are further assessed with an analysis of their correlations. Case-specific analysis includes an NTSB-reported incident as well as one that includes the movement of the ASOS location within the period of analysis.

3.1. Unfiltered Average Gust Factors by Direction

The DFW East Air Traffic Control Tower is approximately 290 m east of the ASOS, which also has the main terminal areas to the west. KGLS, which is less developed than KDFW (Table 2), has a VOR located about 345 m to the west/southwest of the ASOS. KPAE has a large nearby fence-like structure 80 m to the southeast of the ASOS. The average GF_met is relatively high (Figure 2) from these directions at each airport, particularly at KGLS and KPAE, because of the obstructions. KABQ has a grouping of hangars and other buildings to the southwest of the ASOS and buildings to the northeast that account for the more subtle increases in GF_met. The lowest values of GF_met at each airport are in directions where the upwind flow has a significant fetch along a runway. A full listing of isolated and/or impactful siting issues is included in

Table A2. Isolated obstructions and/or other impactful siting issues.

ICAO Airport Code	Isolated Obstructions or Siting Issues
KABQ	None noted
KAOO	House 90 m northeast of ASOS
KAZO	None noted
KBGD	Hangars 50 m northwest of ASOS
KBUF	Airport structure 75 m north–northeast of ASOS
KCKV	House 160 m south–southeast of ASOS
KDFW	ATC tower 290 m east of ASOS
KGEZ	Farm silos 260 m north–northeast of ASOS
KGLS	VOR 345 m west–southwest of ASOS
KHGR	Airport terminal 115 m south–southwest of ASOS
KHSV	Patch of trees 115 m west of ASOS
KIDA	Glide slope equipment 15 m northeast of ASOS
KLXV	ASOS located next to large area of trees
KMHT	Buildings 165 m east–northeast of ASOS
KONT	None noted
KOVE	Road with trees 20 m west of ASOS
KPAE	Large airport structure 80 m southeast of ASOS
KPKD	Hangars 25 m northeast of ASOS
KSLC	None noted
KTCC	None noted

. Figures depicting GF_met at the remaining locations are included in Appendix B.

The GF_avn is quite different, directionally, at each airport compared to the GF_met (Figure 3). KABQ shows a relatively large increase in GF_avn directly to the east, associated with gap winds that flow from Tijeras Canyon [18] with the mouth of the canyon about 12 km to the east, which produces a larger footprint of meteorologically forced gustiness. KGLS has generally higher and more variable GF_avn with winds from the northwest and lower and less variable by direction GF_avn with winds from the southeast, which are assumed to be related to the orientation of the coastline, with the higher gustiness associated with wind coming off the land mass and the lower gustiness from the oceanic directions. The highest GF_avn values at the other three locations are consistent with prevailing wind patterns. Elevated GF_avn values in the directions with nearby obstructions are noted for KGLS and KPAE, which may indicate that local obstructions can be severe enough to be notable in GF_avn. Figures depicting GF_avn at the remaining locations are included in Appendix C. Other sites exhibit a non-local gust footprint, such as KONT, where the GF_avn has higher values from the northeast, which is likely due to infrequent Santa Ana wind events (Figure A7).

3.2. Filtered Average Gust Factors by Direction

Given that the impact of obstructions near the ASOS units is less noticeable with GF_avn and the overall gustiness patterns are dissimilar between the two gust factors, a sensitivity analysis was performed whereby gusts are filtered using a threshold-based approach. Here, we systematically remove gusts less than 1, 5, 10, 15, 20, and 25 kt and compare directionally averaged values for each case. A threshold value of 1 is equivalent to no filtering, given the wind reporting resolution is 1 kt. A threshold near 15 kt is comparable to the 14 kt minimum gust value in ASOS-generated METAR. The impact of filtering of GF_met by 5 kt or greater produced significant differences compared to the unfiltered estimates, particularly in those upstream directions with developed land cover where mechanical turbulence generation is likely to be significant (Figure 4). For example, KABQ had large reductions in GF_met by as much as 0.2 at 45° and approximately 0.1 for the southwest-to-northwest flow sectors. Although there is overall less impact than at KABQ, KDFW exhibits some reduction in GF_met—with a difference on the order of 0.1 for westerly flow downstream of the terminal buildings. Conversely, some of the higher GF_met values in the direction of prominent obstructions at KDFW, KGLS, and KPAE remain despite the filtering. For some of the directions and increasing thresholds, there are no data, and thus, it is not possible to calculate GF_met. This is indicative of the impact of more severe obstructions that both increase gustiness and can also reduce wind speeds due to wind blocking.

The GF_avn is more sensitive to filtering by gust speed (compare Figure 4 and Figure 5). Given that GF_avn values are not normalized, the values increase as the gust filter value is increased. With the exception of the 5 kt threshold, the impact is more incremental than that of the GF_met, with values increasing systematically as the threshold increases. At KABQ (Figure 5a), the gustiness in association with gap winds from the east is present in all levels of filtering, while the two peaks for southwest flow are damped but still somewhat evident without a filter. The gusts from the gap winds are high enough and frequent enough not to be overwhelmed by lower gusts in the unfiltered average and to be detected when filtered for only higher gust values. At KDFW (Figure 5b), the gustiness associated with the ATC tower to the east becomes more noticeable with increased filtering until gust values stop being reported from that direction. The dropout of gust reports may be attributable to the tower, which, while increasing overall gustiness, also provides a reduction in winds due to blocking effects. This is also supported by the fact that all directions, except winds from 80° and 90°, provide average gust factor values for gust filtering of 25+ kt, and winds from 90° additionally do not provide average gust values for gust filtering of 20+ kt. Thus, while overall gustiness is high from the east, pilots may not be aware of higher gusts when winds are from the east. Additionally, the higher gustiness with east winds, while likely impacting a portion of the nearby runway, is not representative of the whole airport. A similar impact is observed in association with the VOR at KGLS, where both enhanced gustiness and blocking effects are evident. KPAE is more problematic to assess, given the poor siting of the ASOS.

3.3. Correlation Between GF_met and GF_avn as a Function of Gust Filtering

Given the sensitivity of both gust factors with respect to filtering, a systematic comparison of their relationship was assessed using the Pearson correlation coefficient. The correlation was evaluated for each of the filtering thresholds. Results of this testing for KABQ, KDFW, KGLS, and KPAE are shown in Figure 6, with the remaining locations presented in Appendix D. As the gust speed threshold is increased, both the slope and correlation trend become more positive. The peak correlation coefficients for all locations occur with filtering of at least 15 kt or greater, with a total of 17 locations having peak correlation coefficients greater than +0.95. It can easily be shown that when Equation (3) is rearranged, the slope is directly proportional to the mean wind speed and, thus, will increase as the gust threshold increases. Additionally, the intercepts of the best-fit line tend toward zero at most of the locations as the filtering level is increased (unshown).

In the absence of any filtering, five of the locations (KAOO, KDFW, KIDA, KONT, and KTCC) have robust negative correlations (between −0.5 and −0.7). We speculate that the cause of this negative correlation stems from the following scenario. ASOS systems are generally sited near the landing zone of the primary instrumented runway, which is typically well aligned with the climatological winds. The prevailing wind direction is often the direction of larger gusts and, thus, would frequently produce high GF_avn in the directions parallel to the runway for which the ASOS system generally views a long upwind fetch (i.e., along the runway). As an example, Manasseh R. & Middleton [19] note the importance of along-runway gusts for the specific case of Sydney Airport in Australia. GF_met, which responds to the lower roughness along the runway directions, would be, in general, low. In cases where there is higher roughness near the runway, GF_met would be larger with fetches perpendicular to the runway. Thus, for wind directions parallel to the runway, GF_avn (GF_met) may be relatively high (low), while for flow perpendicular to the runway, GF_avn (GF_met) may be relatively low (high). Therefore, as GF_met increases, GF_avn would decrease, producing a negative correlation.

The high correlation between the two versions of the gust factor with higher thresholds of gust filtering might be explained by changes predominantly in either or both gust factors. To help provide insight, the correlation between each of the directionally averaged gust factors of the unfiltered and filtered versions, using the 20+ kt threshold, was used to help assess which gust factor is more impacted by the gust filtering.

Table 3. Pearson correlation coefficient between unfiltered and filtered (20+ KT) directionally averaged gust factors at each ASOS location.

ICAO Airport Code	GFmet	GFavn
KABQ	0.40	0.70
KAOO	0.31	0.20
KAZO	0.86	0.20
KBGD	0.81	0.14
KBUF	0.68	0.34
KCKV	0.89	0.41
KDFW	0.75	−0.25
KGEZ	0.71	0.10
KGLS	0.92	0.47
KHGR	0.72	0.05
KHSV	0.43	0.68
KIDA	0.57	0.15
KLXV	0.91	0.66
KMHT	0.49	0.21
KONT	0.66	0.18
KOVE	0.58	0.72
KPAE	−0.17	0.02
KPKD	0.94	0.43
KSLC	0.45	−0.01
KTCC	−0.16	0.41

includes this correlation for all 20 locations. Overall, the highest correlation values are for GF_met and indicate that its directional variance is less influenced by the gust filtering. GF_avn, on the other hand, generally shows much less correlation between the unfiltered and filtered versions. Thus, the better correlation between the two gust factors appears to be predominantly related to filtering impacts on GF_avn. KDFW, for example, has a high correlation with filtered and unfiltered GF_met, and for which the directional GF_met pattern is similar between unfiltered and filtered versions (Figure 3b). GF_avn at KDFW (Figure 4b), though, shows differing directional patterns of gustiness with increased filtering and shows that the GF_avn becomes more prominent with increased filtering from the direction of the ATC tower to the east. KGLS shows similar impacts to KDFW (Figure 3c and Figure 4c), which also demonstrates that the gust impact from the direction of the VOR becomes much more observable in GF_avn with increased gust filtering. In these examples, the impact of prominent obstructions is evident with higher gust filtering that is skewed toward higher-impact gusts. Important exceptions occur at KABQ and KOVE, where the correlation for GF_avn is significantly higher than for GF_met. Interestingly, KABQ is impacted by canyon winds. KOVE, which is situated in the Northern Sacramento Valley of California, is influenced by the along-valley prevailing flow. Both are suggestive of the influence of orographic winds.

3.4. NTSB Case CEN18LA172 Borger, Texas

The NTSB report for the incident at KBGD on 11 May 2018 at 13:27 UTC, which was classified as a loss of control on ground, included the following text:

During a visual approach to runway 21 (3897 feet by 100 feet, dry asphalt), the pilot stated that he flew an upwind pattern entry and on short final, experienced “strong updrafts from unusual landforms”. The pilot landed the tailwheel equipped airplane on runway 21, near the intersection of runway 17/35 (about 1650 feet down runway 21), and “experienced strong wind gusts causing swerving”.

The report also indicated that the winds were from 220° (southwest) between 23 and 32 kt, which is consistent with archived METAR reports. Topographic features in the vicinity of the airport are visible in Figure 7, including features to the west and southwest of the intersection of the two runways. These are presumed to be the landforms to which the pilot refers. However, the unfiltered GF_met (Figure A1) and GF_avn (Figure A5) do not indicate high values from the southwest. This is not surprising, given that the ASOS is located to the southeast of this topographic feature. The GF_met, though, is relatively high from the northwest (approximately 0.5), which is in the general direction of the most prominent terrain features responsible for producing the mechanical turbulence in this incident. Unfortunately, the high GF_met values observed by the ASOS from the northwest are most likely due to the proximity of hangars that are located only 50 m in that direction. Thus, poor ASOS siting can limit the utility for assessing mechanical turbulence associated with natural obstructions in the vicinity of the airport. Also, more obviously, ASOS provides directional gustiness based on its location, and this directional variance cannot be assumed to apply at other locations within the airport.

3.5. Impact of ASOS Location Change at Manchester–Boston Regional Airport (KMHT)

The ASOS unit at KMHT was relocated as part of airport facilities construction in 2022. The old location was positioned on the west side of runway 17/35, with the terminal to the west, while the new location is positioned further south on the east side of runway 17/35, with major obstructions to the east (Figure 8). An analysis of the gustiness in 2021, before the relocation, and in 2024, after the relocation, was conducted to test the impacts on the two gust factors. GF_met (Figure 9a,c,e) shows a relatively large change between the two years, with peak gustiness shifting from the west to the east and increasing from 0.4 to greater than 0.5. The impact on GF_met is consistent with changes in the location of the ASOS with respect to the obstructions (Figure 9b,d,f). The primary impact on the GF_avn is an increase in the magnitude, with the directional patterns in gustiness largely preserved. Notably, the new location shows similar or increased average aviation and meteorological gust factors (Figure 9d,f) for all directions, with GF_met showing the least amount of change in directions oriented along runway 17/35, which emphasizes the previous discussion of GF_met being generally low along longer fetch directions more parallel to the runways they are sited next to. The larger values of GF_met to the east are likely due to the new location having airport structures and buildings situated very close to the east. The cause of the overall increase in GF_avn is less certain, but perhaps related to the significant addition of airport structures that necessitated the ASOS move. Information regarding the impact of the ASOS relocations on observed gustiness is not readily available to the aviation community.

4. Discussion

Gust factors are used in piloting, in particular, by fixed-wing aircraft on final approach, and large values may be caused by mechanical turbulence produced by natural and/or human-made obstructions. Meteorological gust factors are often used to estimate surface roughness conditions for providing standardized wind speed and gusts, for example, in tropical cyclones. Meteorological (GF_met) and aviation gust factors (GF_avn) were analyzed from six years of one-minute ASOS observations at 20 airports, including small, medium, large, and non-hub facilities. A modified version of the meteorological gust factor is used and is shown to be equivalent to the aviation gust factor normalized by the sustained wind speed. The recalibration of GF_met facilitates the comparison between the two gust factors since the minimum values for both are zero when the gust speed equals the sustained wind speed. The key findings, which are summarized below, are broken into two categories that were the focus of this assessment: gust factor attribution and gust factor representativeness, which are not mutually exclusive. These include the following:

4.1. Gust Factor Attribution

• In the absence of filtering for gust speed, both GF_met and GF_avn are often poorly correlated. When considering unfiltered gust factors, GF_met is sensitive to surface roughness from nearby structures, which can be imperceptible in GF_avn. Conversely, GF_avn is shown to be responsive to meteorological features, such as canyon winds, which have a muted response in GF_met. Both gust factors would be important to address when considering impacts on gustiness due to mechanical turbulence by features of different spatial scales;
• GF_met and GF_avn are, in general, highly correlated when gusts less than 15 knots are excluded from the analysis. The higher correlation is, in general, due to the filtered GF_avn becoming more responsive to the gustiness generated by nearby features. Such filtering may be useful for highlighting the impacts of higher gustiness. Aircraft-specific gust load information may need to be considered when applying such filters;
• Impacts of nearby orography or airport structures, as noted in reports from pilots, can be difficult to evaluate from the given ASOS information due to the localized nature of the gustiness relative to the ASOS location. Additionally, there are masking issues related to the gustiness caused by structures in proximity to the ASOS, especially when the system is poorly sited.

4.2. Gust Factor Representativeness

• GF_met is sensitive to local obstructions ranging from small nearby airport equipment and small buildings to larger generalized areas of mixed airport-related development. While the unfiltered GF_avn is less sensitive to local obstructions, there are cases of impactful obstructions that can mask higher gusts observed by aviators than are reported by the ASOS. The directional representativeness or quality of the gust information could be provided to aviators as a caveat for better situational awareness. Databases of directionally dependent surface roughness at ASOS locations, such as those by Masters et al. [20], have been created in the past and reflect the importance of site-specific guidance;
• A subset of airport locations exhibits negative correlations (i.e., GF_avn decreases as GF_met increases). This may be caused by ASOS being sited where development and/or natural roughness elements near a runway are aligned with the prevailing flow;
• An assessment of an ASOS relocation indicated that directional variation in GF_met was impacted (altering the peak gust direction from west to east) while being largely preserved for GF_avn; however, the magnitude increased in most directions. This indicates that the unfiltered GF_avn may be more representative of the larger-scale roughness and/or meteorologically driven gust events. Information regarding the impacts of ASOS location changes is not readily available to the aviation community.

Local obstructions in the vicinity of the ASOS can make the evaluation of mechanical turbulence by natural features difficult. The approach presented here is important with respect to the interpretation of the climatological gustiness at aerodromes. This type of analysis is necessary to properly advise aviators, assign attribution of gusts for specific aviation incidents, and assess ASOS equipment siting issues. Providing the user community with site-specific metadata regarding the representativeness and/or quality of the gust information, by direction, is recommended.