Slowest Start to East PACific Season since 1977
as of 0806 or 6 August 2024
Update as of 0826 or 26 August 2024
I calculate Year-To-Date (YTD) TC activity every day in both hemispheres and make three kinds of plots for the current 2024 NHEM season:
- Spectographs (https://tcact.wxmap2.com/cur/spec.htm) -- visualize TC by intensity in all basins
- Maps (https://tcact.wxmap2.com/cur/ll.htm) -- lat-lon maps showing TC activity and anomalies
- Time-Series (https://tcact.wxmap2.com/cur/ts.htm) -- time series of monthly means and climatology
The basins I follow in NHEM are:
- EPAC -- Eastern north PACific (includes the central PAC)
- WPAC -- Western north PACific
- LANT -- north atLANTic
- IO -- north Indian Ocean (both the Bay of Bengal and the Arabian Sea)
My TC activity metrics are slightly different than ACE (Accumulated Cyclone Energy) and are calculated YTD (units in [] Vmax is the NHC/JTWC intensity) and for the entire season (yearly):
- sTCd [days]-- scaled TC days accumulate every 6 h
- TC = ( (Vmax * 6 h) / Vmax ) * (24 h /day)
- sACEd [days] -- scaled ACE days accumulate every 6 h:
- when Vmax >= 35 kts ACE = ((Vmax*Vmax*6h)/(35*35kts)*(24h/d)
- sHUACEd [days] -- scaled HUrricane ACE days accumulate every 6 h when Vmax >= 65 kts ACE = ( (Vmax*Vmax*6h)/(65*65kts) * (24h/d)
ACE is typically expressed in Tjoules but I prefer units that are more natural to me, i.e., why I scale to make the units days. Otherwise, my sACEd is identical to the standard ACE.
sTCd is my original time-intensity metric that avoids sensitivity to the accuracy of Vmax ... why I'm not a big fan of Kerry Emanuel's power metric. Regardless, ACE is the de facto standard.
The anomaly is expressed as a percent departure from a 30-y climatology 19990101-20201231. For the YTD anomalies, the climatology is calculated from the beginning of the season 0101 to the YTD. In this blog 0101-0806 or 1 January - 6 August. This is the 0101-0826 update.
At the end of the season (YYYY1231 for NHEM and YYYY0630 for SHEM) the YTD TC activity is the total activity for the year.
I compare the YTD (20240806) activity (1977-2024) to the annual (1976-2023) to show how early-season anomalies differ from those of the full season. The anomaly unit is % and the metric is sACEd. For example, if the actual sACEd is 30 d and the climatology is 15 d, the anomaly would be 100%
YTD EPAC v LANT as of 20240806 -- 0826 update
Looking at my maps every day, I noticed that as of 6 August 2024 the EPAC YTD sACEd anomaly was very, very low at -93 % as of 26 August is -49%. I did some quick python and matplotlib to see if it was lower in the past and found this current value is a record going back to 1977. NB: the obs in EPAC were not very good in the late 70, 80s. The python was updated today to plot two basins and calculate statistics.
 |
| Fig. 1 YTD (0806) anomalies EPAC v LANT |
The 0826 update:
 |
| Fig. 1a YTD (0826) anomalies EPAC v LANT |
In contrast to the EPAC, the LANT is above normal, but not a YTD record for 0806. Note the negative correlation of -0.33 between EPAC and LANT anomalies.
Current YTD anomalies as of 0826:
EPAC: -49%
LANT: +83%
Only 2005 LANT season had a faster start...with a record anomaly of 491%!
The LANT has slowed down considerably since 0806
There is a noticeable trend in the LANT (orange line) for the 2005 -- present era -- more early-season activity. My guess is that NHC is calling early storms (often subtropicals) more often than before 2006 because of better satellite data. Coincidentally I worked at NHC 2006-2008 as the lead Tech Dev...
YTD EPAC v WPAC as of 20240806 -- 0826 update
I also noticed that the NHEM activity was low as well -- because WPAC was very low as shown below comparing YTD EPAC v WPAC:
 |
| Fig.2 YTD anomalies EPAC v WPAC |
The 0826 update:
 |
| Fig.2a YTD anomalies EPAC v WPAC as of 0826 |
The WPAC season is still below normal at -59% as of 0826 v -69% as of 0806
YTD EPAC v LANT v WPAC as of 20240826 -- 0806 v 0826 gif
Here are gif versions comparing 0806 to 0826
 |
| Fig.1b YTD anomalies EPAC v LANT -- 0806 v 0826 |
 |
| Fig.2b YTD anomalies EPAC v WPAC -- 0806 v 0826 |
Season EPAC v LANT v WPAC for 1976-2023
Here are the same plots as in Figs. 1 & 2 but now for the entire season.
 |
| Fig. 3. As in Fig. 1a but for the annual/season anomalies |
 |
| Fig. 4. As in Fig. 2a but for the annual/season anomalies |
Summary of YTD v season - EPAC v LANT
Here is a summary of the stats:
EPAC v LANT Year-To-Date (0806) 1977-2024:
Corr: -0.33
EPAC CUR: -93% min: -93% in 2024; max: 186% in 1978
LANT CUR +350%. min: -97% in 1988; max: +491% in 2005
EPAC v LANT Year-To-Date (0826) 1977-2024:
Corr: -0.33
EPAC CUR: -49% min: -93% in 2024; max: 186% in 1978
LANT CUR: +85% min: -97% in 1988; max: +491% in 2005
EPAC v LANT YEARly 1976-2023:
Corr: -0.37
EPAC min: -83% in 1977; max: +139% in 2018
LANT min: -84% in 1983; max +132% in 2005
Note the negative correlation of ~ -0.3 between EPAC and LANT anomalies for both YTD and seasonally.
Summary of YTD v season - EPAC v WPAC
In the meantime, WPAC is off to a slow start as well as of 0806 (-69% v -93% in EPAC).
EPAC v WPAC Year-To-Date (0806) 1977-2024:
Corr: -0.02
EPAC CUR: -93% min: -93% in 2024; max: +186% in 1978
WPAC CUR: -75% min: -92% in 1998; max: +166% in 2015
EPAC v WPAC Year-To-Date (0826) 1977-2024:
Corr: +0.18
EPAC CUR: -49% min: -93% in 2024; max: +186% in 1978
WPAC CUR: -59% min: -92% in 1998; max: +166% in 2015
EPAC v WPAC YEARly 1976-2023:
Corr: +0.42
EPAC min: -83% in 1977; max: +139% in 2018
WPAC min: -65% in 1999; max: +124% in 1997
On a yearly basis, WPAC/EPAC anomalies are relatively well correlated at +0.42, but not YTD for 0806. As of 0826, however, the correlation is now positive at 0.18.
Generally, August WPAC activity is ahead of EPAC/LANT (putting on my JTWC TDO hat)
Conclusions
This curious early season activity is just that. We're moving into the peak of the NHEM season and there will be changes in TC activity, but what's happening in EPAC (the season peaks earlier than in the LANT) will be interesting to follow...
As of 0826, the LANT activity essentially stopped. The old rule of thumb is that there is always a storm in the LANT on Labor Day. If the current (0826) NHC 7-d forecast verifies, then this year's Labor Day (2 September) would be one of the rare times there is no storm...
