Some Applications of Indices to Forecasting

1
Some Applications of Indices to Forecasting

• 12th Great Divide Workshop, 10/7/2008
• Matthew J. Bunkers, SOO Rapid City, SD

2
Outline

• Make note of several indices
• Discuss utility attributes of indices ( / -)
• Show several examples of testing indices for
  operations implications for training

3
A cornucopia of indices
Lifted Index (LI)
Total Totals (TT)
K Index (KI)
Showalter Index (SI)
Severe Weather Threat (SWEAT)
CAP Strength (700 mb LI)
Lapse Rate (LR)
Relative Humidity (RH)
Can be calculated over many different
layers/levels/parcels
4
A cornucopia of indices
Lifted Index (LI)
Total Totals (TT)
K Index (KI)
Showalter Index (SI)
Severe Weather Threat (SWEAT)
CAP Strength (700 mb LI)
Lapse Rate (LR)
Relative Humidity (RH)
Lifted Condensation Level (LCL)
Level of Free Convection (LFC)
Equilibrium Level (EL)
Wet Bulb Zero (WBZ)
Melting Level (MLT)
Warm Cloud Depth (WCD)
Precipitable Water (PW)
Equivalent Potential Temperature (?e)
Moisture Flux Convergence (MFC)
Can be calculated over many different
layers/levels/parcels
5
A cornucopia of indices
Convective Available Potential Energy (CAPE)
Lifted Index (LI)
Convective Inhibition (CIN)
Total Totals (TT)
Bulk Richardson Number (BRN)
K Index (KI)
Bulk Richardson Number Shear (BRNSHR)
Showalter Index (SI)
Bulk Vertical Wind Shear
Severe Weather Threat (SWEAT)
Total Vertical Wind Shear
CAP Strength (700 mb LI)
Storm-Relative Wind
Lapse Rate (LR)
Storm-Relative Helicity (SRH)
Relative Humidity (RH)
Downdraft CAPE (DCAPE)
Lifted Condensation Level (LCL)
Normalized CAPE (nCAPE)
Level of Free Convection (LFC)
Equilibrium Level (EL)
Wet Bulb Zero (WBZ)
Melting Level (MLT)
Warm Cloud Depth (WCD)
Precipitable Water (PW)
Equivalent Potential Temperature (?e)
Moisture Flux Convergence (MFC)
Can be calculated over many different
layers/levels/parcels
Can be calculated over many different
layers/levels/parcels
6
A cornucopia of indices
Convective Available Potential Energy (CAPE)
Lifted Index (LI)
Convective Inhibition (CIN)
Total Totals (TT)
Bulk Richardson Number (BRN)
K Index (KI)
Bulk Richardson Number Shear (BRNSHR)
Showalter Index (SI)
Bulk Vertical Wind Shear
Severe Weather Threat (SWEAT)
Total Vertical Wind Shear
CAP Strength (700 mb LI)
Storm-Relative Wind
Lapse Rate (LR)
Storm-Relative Helicity (SRH)
Relative Humidity (RH)
Downdraft CAPE (DCAPE)
Lifted Condensation Level (LCL)
Normalized CAPE (nCAPE)
Level of Free Convection (LFC)
Wind Index (WINDEX)
Equilibrium Level (EL)
Dry Microburst Index (DMI)
Wet Bulb Zero (WBZ)
Theta-E Index (TEI)
Melting Level (MLT)
Microburst Day Potential Index (MDPI)
Warm Cloud Depth (WCD)
Wet Microburst Severity Index (WMSI)
Precipitable Water (PW)
Equivalent Potential Temperature (?e)
Moisture Flux Convergence (MFC)
Can be calculated over many different
layers/levels/parcels
Can be calculated over many different
layers/levels/parcels
7
A cornucopia of indices
Convective Available Potential Energy (CAPE)
Lifted Index (LI)
HI Haines Index
Convective Inhibition (CIN)
Total Totals (TT)
HMI Hybrid Microburst Index
Bulk Richardson Number (BRN)
K Index (KI)
LSI Lid Strength Index
DCI Deep Convective Index
Bulk Richardson Number Shear (BRNSHR)
Showalter Index (SI)
TQ Index for low-topped instability
Bulk Vertical Wind Shear
Severe Weather Threat (SWEAT)
Total Vertical Wind Shear
CAP Strength (700 mb LI)
Storm-Relative Wind
Lapse Rate (LR)
Storm-Relative Helicity (SRH)
Relative Humidity (RH)
Downdraft CAPE (DCAPE)
Lifted Condensation Level (LCL)
Normalized CAPE (nCAPE)
Level of Free Convection (LFC)
Wind Index (WINDEX)
Equilibrium Level (EL)
Dry Microburst Index (DMI)
Wet Bulb Zero (WBZ)
Theta-E Index (TEI)
Melting Level (MLT)
Microburst Day Potential Index (MDPI)
Warm Cloud Depth (WCD)
Wet Microburst Severity Index (WMSI)
Precipitable Water (PW)
Equivalent Potential Temperature (?e)
Moisture Flux Convergence (MFC)
Can be calculated over many different
layers/levels/parcels
8
A cornucopia of indices
Convective Available Potential Energy (CAPE)
Lifted Index (LI)
HI Haines Index
Convective Inhibition (CIN)
Total Totals (TT)
HMI Hybrid Microburst Index
Bulk Richardson Number (BRN)
K Index (KI)
LSI Lid Strength Index
DCI Deep Convective Index
Bulk Richardson Number Shear (BRNSHR)
Showalter Index (SI)
TQ Index for low-topped instability
Bulk Vertical Wind Shear
Severe Weather Threat (SWEAT)
Indices of Indices (Inbreeding)
Total Vertical Wind Shear
CAP Strength (700 mb LI)
Energy-Helicity Index (EHI)
Storm-Relative Wind
Lapse Rate (LR)
Storm-Relative Helicity (SRH)
Vorticity Generation Parameter (VGP)
Relative Humidity (RH)
Supercell Composite Parameter (SCP)
Downdraft CAPE (DCAPE)
Lifted Condensation Level (LCL)
Significant Tornado Parameter (STP)
Normalized CAPE (nCAPE)
Level of Free Convection (LFC)
Significant Hail Parameter (SHIP)
Wind Index (WINDEX)
Equilibrium Level (EL)
Significant Severe Parameter (SSP)
Dry Microburst Index (DMI)
Wet Bulb Zero (WBZ)
Strong Tornado Parameter (STP)
Theta-E Index (TEI)
Melting Level (MLT)
Microburst Day Potential Index (MDPI)
Warm Cloud Depth (WCD)
Wet Microburst Severity Index (WMSI)
Precipitable Water (PW)
Equivalent Potential Temperature (?e)
Moisture Flux Convergence (MFC)
Can be calculated over many different
layers/levels/parcels
9
A cornucopia of indices
Convective Available Potential Energy (CAPE)
Lifted Index (LI)
Total Totals (TT)
Convective Inhibition (CIN)
HI Haines Index
HMI Hybrid Microburst Index
K Index (KI)
Bulk Richardson Number (BRN)
LSI Lid Strength Index
DCI Deep Convective Index
Showalter Index (SI)
Bulk Richardson Number Shear (BRNSHR)
TQ Index for low-topped instability
Severe Weather Threat (SWEAT)
Bulk Vertical Wind Shear
CAP Strength (700 mb LI)
Total Vertical Wind Shear
Indices of Indices (Inbreeding)
Lapse Rate (LR)
Storm-Relative Wind
Energy-Helicity Index (EHI)
Storm-Relative Helicity (SRH)
Vorticity Generation Parameter (VGP)
Relative Humidity (RH)
Vorticity Generation Parameter (VGP)
Supercell Composite Parameter (SCP)
Downdraft CAPE (DCAPE)
Supercell Composite Parameter (SCP)
Lifted Condensation Level (LCL)
Normalized CAPE (nCAPE)
Significant Tornado Parameter (STP)
Level of Free Convection (LFC)
Significant Hail Parameter (SHIP)
Equilibrium Level (EL)
Wind Index (WINDEX)
Significant Severe Parameter (SSP)
Wet Bulb Zero (WBZ)
Dry Microburst Index (DMI)
Strong Tornado Parameter (STP)
Melting Level (MLT)
Theta-E Index (TEI)
Warm Cloud Depth (WCD)
Microburst Day Potential Index (MDPI)
Precipitable Water (PW)
Wet Microburst Severity Index (WMSI)
Equivalent Potential Temperature (?e)
Mesoscale Convective System Forecast Index (MCS
Index) ? a recent index published in WAF (2007)
Moisture Flux Convergence (MFC)
? This list is not nearly exhaustive!
Can be calculated over many different
layers/levels/parcels
10
Whats a forecaster to do?
11
Outline

• Make note of several indices
• Discuss utility attributes of indices ( / -)
• Show several examples of testing indices for
  operations implications for training

12
Attributes of indices

• Doswell and Schultz (2006)
• On the Use of Indices and Parameters in
  Forecasting Severe Storms
• Electronic Journal of Severe Storms Meteorology
• http//www.ejssm.org/

13
Benefits of indices

• Can summarize large amounts of data
• Can quickly draw attention to critical areas
  for further diagnosis
• Both are attractive when under time pressure

14
Index limitations

• Not necessarily forecast parameters may be
  diagnostic (e.g., SPC meso page)
• Diagnostic variables give current state(?
  ??/?t), where ? STP, SCP, CAPE, etc.
• Most indices are not rigorously developed or
  validated arbitrarily combined variables

15
Index limitations

• Can lead to faulty perceptions of atmosphere via
  over-simplification
• Little value in isolation different combos can
  produce similar values
• Flavor of the parameter? (e.g., EHI and its
  inputs)
• Constituents can evolve quasi-independently
• Action often occurs at The Edge next three
  slides

16
The Edge 20 Jun 2006 Rushville, NE
Tornadic left-moving supercell
?
LSCP
(1-EF1)
17
The Edge 16 Sep 2006 Rogers, MN
(1-EF2)
18
The Edge 28 Feb 2007 Eastern KS
(1-EF4)
Important to train new forecasters not to focus
on bulls-eyes.
19
Outline

• Make note of several indices
• Discuss utility attributes of indices ( / -)
• Show several examples of testing indices for
  operations implications for training

20
Example of testing an index

• Jirak and Cotton (2007, WAF)
• MCS index for conditional development of MCSs
• Function of best LI, 0-3km shear, and 700mb
  Tadv
• Convert terms to standard normal and summed
• Appears to be physically based
• Developed using NARR data
• Tested at WFO Rapid City and found problems
• Operational datasets produced different results
• Didnt implement at our office

21
MCS index on WES
Image MCS index White lines 700mb Tadv MCS
index and 700mb Tadv looked very similar Was
verified with 20 cases
22
MCS index testing

• Let LI vary from 3 to -12 (?C)
• Let shear vary from 0 to 25 (m s-1)
• Let Tadv vary from -1 to 2 (C hr-1)

Mean
Std Dev
Means and standard deviations based on NARR
dataset (JC07).
23
Three terms of MCS index
Using JC07s equation and reasonable ranges for
the 3 terms Ideally all 3 lines should be the
same
24
MCS index summary

• LI component ? -2 to 2
• Shear component ? -2 to 2
• Tadv component ? -4 to 6 (should be -2 to 2)
• NARR data underrepresented the operational Tadv
  range
• MCS index basically proxy for 700mb Tadv
• Conclusion not suitable for operations (authors
  updating to use stddev of range)

25
Outline

• Make note of several indices
• Discuss utility attributes of indices ( / -)
• Show several examples of testing indices for
  operations implications for training

26
The STP index

• Thompson et al. (2003, WAF)
• Significant Tornado Parameter (STP)
• Mean-layer CAPE (MLCAPE, lowest 100mb)
• 0-6km shear vector magnitude (SHR6)
• 0-1km storm-relative helicity (SRH1)
• Mean layer LCL (MLLCL, lowest 100mb)

27
Lets test this

• Estimate valid ranges and calculate each term
• For example MLCAPE 100 to 5000 J kg-1
• Term 1 thus ranges from 0.1 to 5
• (100/1000) 0.1
• (5000/1000) 5

28
Versions of the STP
29
Versions of the STP
30
Versions of the STP
If you use them, know your indices!
31
Outline

• Make note of several indices
• Discuss utility attributes of indices ( / -)
• Show several examples of testing indices for
  operations implications for training

32
Supercell Composite Parameter (SCP)

• Function of MUCAPE, Eff. Shear, and Eff. SRH
• Can run similar tests as for the STP
• Testing suggests SCP can be misleading

33
SCP Potential Pitfall
Dominant right-mover
34
Modeling Results
SRH (RM) 62 SRH (LM) -226
Hodograph didnt turn enough in lowest 3?6 km to
strongly favor LM.
Environmental heterogeneity and low-level wind
variability also factors.
35
Outline

• Make note of several indices
• Discuss utility attributes of indices ( / -)
• Show several examples of testing indices for
  operations implications for training

36
Example of coord system sensitivity

• SWEAT Index (SW) 27-28 May 2001 OK case
• SW 331
• What if 850 wspd 15 kts? (SW 429)
• Now what if wdir 30?to left (SW 331)

37
Supercell motion example BUFKIT

• Bunkers et al. (2000)
• Non-weighted MW for supercell motion, every 500
  meters
• BUFKIT
• Uses ALL data for MW produces low-level bias
• Supercell motion often too slowso beware of
  BUFKIT algorithm!

38
SCM Excel vs. BUFKIT
275 28 kts
12-kt difference between the two! but
AWIPS is okay
39
Summary for indices

• Look at the raw data (e.g., surface maps,
  soundings, 0-1km shear, MLLCL, etc.)
• View the indices constituent components (e.g.,
  4-panel mode)STP 2 means what?
• Test new indices before implementing them in
  operations (e.g., the MCS index)
• Folly to develop indices away from operations

40
One final thought

• The authors most regrettable severe storm
  forecast mistakes have arisen from ignoring data
  that were relevant to the daily diagnosisand/or
  failing to complete the diagnosis on what
  initially appeared to be a benign weather day.
• Al Moller (2001, Severe Convective Storms
  Monograph)
• Analysis and diagnosis of observational data is
  critical yet this has become a lost art.