ARTMIP Algorithms

Listed algorithms (A1, A2, etc.) have submitted catalogues for Tier 1. Algorithm numbering is determined by overall alphabetical order with the first set of numbers assigned to developers who participated in the 1-month proof of concept (A1 – A15), see GMD paper, followed by Tier 1-only participation (A16-27). ARTMIP algorithm identifier will remain consistent across tiers and scientific papers.

Algorithm Grid

Algorithm Grid, by Christine Shields

Developer Type Geometry Req. Threshold Req. Temporal Req. Region DIO/Reference
A16,A23,A24 Brands et al Condition > 1500 km Both Relative and Absolute IVT:
95th(A16)/90th(A23)/90th(A24) percentile at point of detection, 90th/85th/85th percentile along the AR structure, consider all months for threshold calculation with a min threshold of 240/500/250 kgm-1s-1.
Spatial tracking guided by vector IVT.
Time slice 150°W to 30°E
30°N to 62°N
A1 Gershunov et alb Condition and Track >= 1500km long Absolute: 250kgm-1s-1 IVT
1.5cm IWV
Time stitching
-18 hours (3 time steps for 6 hourly data)
Western U.S. 10.1002/2017GL074175
A2 Goldensonb,d Condition > 2000km long and < 1000km wide
Object recognition
Absolute: 2cm IWV Time slice Western U.S. 10.1175/JCLI-D-18-0268.1
A17 Gorodetskaya
et al
Condition IWV > thresh. at the coast (within defined longitudinal sector) and continuously at all latitudes for ≥ 20° equatorward (length > 2000 km), within ± 15° longitude sector (width of 30° ~ 1000 km at 70°S; requirement of meridional extent) Relative: aZN using IWV adjusted for reduced tropospheric moisture holding capacity at low temperatures
(ARcoeff = 0.2)
Time slice Polar (East Antarctica) 10.1002/2014GL060881
A3 Guan and Waliserbc Condition Length >2000km and length width ratio >2; Coherent IVT direction within 45° of AR shape orientation and with a poleward component Relative: 85th percentile IVT; Absolute min requirement designed for polar locations: 100kgm-1s-1 IVT Time slice Global 10.1002/2015JD024257
A4 Hagos et al. b Condition Dependent on threshold requirements to determine footprint;
> 2000 km long and < 1000 km wide
Absolute: 2cm IWV 10ms-1 wind speed Time slice Western U.S. 10.1175/JCLI-D-14-00567.1
A27 Krinitskiy et al. Condition Geometry reqirements Length >= 250km; Length-to-width >=5; Length is estimated along the "ridge" taking IVT into account. Width is the median of widths estimated in each point of AR ridge. Relative: IVT-IVT_RM >_ 100. IVT_RM is climatological IVT running mean with 20-day windows Time slice Global Experimental
A18 Lavers et al. Condition 4.5° latitude movement allowed Relative: ~85th percentile determined by evaluation of reanalysis products Time slice UK, Western US 10.1029/2012JD018027
A5 Leung and Qianb Track Moisture flux has an eastward or northward component at landfall; tracks originating north of 25N and east of 140W are rejected Absolute: mean IVT along track > 500 kgm-1s-1 and IVT at landfall > 200 kgm-1s-1; grid points up to 500km to the north and south along the AR tracks are included as part of the AR if their mean IVT > 300 kgm-1s-1 Time slice Western U.S. 10.1029/2008GL036445
A6,A7 Lora et al.b Condition Length >= 2000km Relative: IVT 100kgm-1s-1 above climatological area means for N. Pacific Time slice Global (A6)
North Pacific (A7)
A25,A26 Mahesh and O'Brien (CASCADE) Condition Convolutional neural network to replicate ARTMIP mean Threshold-free; IVT(A25) and IWV (A26) is the input, output is probability of AR presence Time slice Global Experimental
Mahoney et al. Condition and Track Length >= 1500km, Width <=1500km Absolute: ARDT-IVT 500kgm-1s-1 for SEUS. See Wick Southeast U.S. 10.1175/MWR-D-15-0279.1
(uses Wick)
A19 Mundhenk Condition >1400km length, aspect ratio 1:4, lat limit >16N/S, axis orientation based on IVT Relative IVT percentiles and/or anomalies both temporal and spatial Time slice Global 10.1175/JCLI-D-15-0655.1
A20 Muszynski et al. Condition Topological analysis and machine learned Threshold-free N/A Western U.S., adaptable to other regions Experimental
A8 Payne and Magnusdottirbc Condition Length > 1200km, landfalling only Relative: 85th Percentile of maximum IVT (1000-500mb) Absolute: IWV >2cm, 850mb wind speed > 10m/s Time stitching (12-hour minimum) Western U.S. 10.1002/2015JD023586
Ralph et al. Condition Length >= 2000km, Width <= 1000km Absolute: IWV 2cm Time slice Western U.S. 10.1175/1520-0493(2004)132<1721: sacaoo>;2
A9 Ramos et al.bc; Blamey et al.* Condition Detected for reference meridians, length >=1500km (1800km*), latitudinal movement <4.5°N Relative: IVT 85th percentile (1000-300mb) Time slice, but 18-hour minimum for persistent ARs Western Europe, South Africa*, adaptable to other regions 10.5194/esd-7-371-2016
A10 Rutz et al.b Condition Length >= 2000km Absolute: IVT (surface to 100mb) = 250kgm-1s-1 Time slice Global, low value on tropics 10.1175/MWR-D-13-00168.1
Shearer et al. (AR-CONNECT,"ARC" ) Track Object identification; Average IVT direction of pixels in a discrete 2-D AR slice must be within 30° of the AR slice’s semi-major axis Absolute: IVT = 300 kgm-1s-1 Time stitching, minimum 24-hour period Global
A11, A12, A13 Sellars et al.b (CONNECT) Track Object identification Absolute: IVT, thresholds tested = 300 (A11), 500 (A12), 700 (A13) kgm-1s-1 Time stitching, minimum 24-hour period Global 10.1002/2013EO320001
A14 Shields and Kiehlb Condition Ratio 2:1, length to width grid points min 200km length; 850mb wind direction from specified regional quadrants, landfalling only Relative: aZN moisture threshold using IWV; Wind threshold defined by regional 85th percentile 850mb wind magnitudes Time slice Western U.S. Iberian Peninsula, UK, adaptable but regional specific 10.1002/2016GL069476
A15 TEMPESTb Track Laplacian IVT thresholds most effective for widths >1000km; cluster size minimum = 120000km2 IVT >=250kgm-1s-1 Time stitching Global, but latitude >=15° Experimental
A21 Viale et al. Condition Length >2000km (>1500km north of 35°S) and length-width ratio>2; Coherent IVT direction within 45° of AR mean direction and with a poleward component Relative: 85th percentile IVT;
Absolute min requirement designed for subtropical locations: 100kgm-1s-1 IVT; a frontal zone intercept or locate south (up to 50 km), i.e., mag. of the horizontal gradient of the 1000-850hPa thickness > 5 m 100 km-1)
Time slice Southwestern South America HM-D-18-0006
Viale et al., 2018, JHM, submitted
A22 Walton et al. Condition and Track Length >= 2000 km Relative: IVT > 250 kg/m/s + daily IVT climatology Time stitching, minimum 24-hour period Western U.S. Experimental
Wick et al. Condition and Track >=2000km long, <= 1000km wide object identification involving shape and axis Absolute: ARDT-IWV >2cm Time slice and stitching Regional 10.1109/TGRS.2012.2211024

aZN relative threshold formula: Q >= Qzonal_mean + ARcoeff (Qzonalmax – Qzonamean) where Q = moisture variable, either IVT (kg m-1s-1) or IWV(cm). ARcoeff = 0.3 except where noted (Zhu and Newell, 1998). The Gorodetskaya method uses Qsat, where Qsat represents maximum moisture holding capacity calculated based on temperature (Clausius-Clapeyron), an important distinction for polar ARs. Additional analysis on the ZN method can be found in Newman et al., 201

bMethods used in a 1-month proof-of-concept test (Section 5). These methods are assigned an algorithm id, i.e. A1, A2, etc.

cThese 1-month proof-of-concept methods apply a percentile approach to determining ARs. A3 and A8 applied the full MERRA2 climatology to compute percentiles. A9, applied the Feb 2017 climatology for this test only. For the full catalogues, A9 will apply extended winter and extended summer climatologies to compute percentiles. Please refer to individual publications (DOI reference column in this table) for climatologies used in earlier published studies by each developer. The climatology used to compute percentile is often dependent on the dataset (re-analysis or model data) being used.

dVersion 1.1 fixes an error in Version 1.0 that resulted in skipping some random time slices.