GaWx

At 2:30 PM, it is windy from the NW with light rain and 57 F due to the nearby offshore low. This is ~27 F BN for mid afternoon on 5/27! Just 4 years ago at this time on this date it was ~100 F, 43 warmer! From this area to Hunter, we're at ~2.25" of rain today, ~2" more than expected as of yesterday.

This area has today gotten much more rain than expected yesterday with ~2" already fallen so far vs <1/2" progged then for the entire day! We even had thunder recently, which hadn't been expected. Temps have been in the chilly mid 50s with N wind gusts into the 30s. The 24 hour high is likely going to be the 12:01 AM's 63, which would be 5 below the 1972 record low high of 68! Normal high is 87. The walks the last few evenings have been fabulous for this time of year with dewpoints in the upper 40s (last evening) to upper 50s!

I'd like to look at the BoM model's May forecast's bias more closely, since it has been discussed quite a bit recently due to its extremely warm May forecast, to see if it really hasn't had a Eurolike strong warm bias when averaged out like I had been thinking. The BoM has been using "Access" since 2020 and POAMA prior to that. Year: May forecast for ASO BoM/Euro, actual ASO, BoM miss vs Euro miss 2022: -0.4/-0.3 -1.0 +0.6 vs +0.7 2021: -0.5/0.0 -0.7 +0.2 vs +0.7 2020: -0.8/-0.3 -0.9 +0.1 vs +0.6 2019: +0.2/+0.7 +0.2 0.0 vs +0.5 2018: -0.2/+0.6 +0.5 -0.7 vs +0.1 2017: +0.2/+0.8. -0.4 +0.6 vs +1.2 2016: -0.8/-0.5 -0.6 -0.2 vs +0.1 2015: +1.6/+2.4 +2.2 -0.6 vs +0.2 2014: +0.8/+1.5 +0.2 +0.6 vs +1.3 2013: -0.8/-0.1 -0.3. -0.5 vs +0.2 ----------------------------------- 2012: +0.9/+0.8 +0.4 +0.5 vs +0.4 2011: +1.2/-0.2 -0.8 +2.0 vs +0.6 2010: -0.4/-0.9 -1.6 +1.2 vs +0.7 2009: +1.4/+0.8 +0.7 +0.7 vs +0.1 2008: +0.3/-0.2 -0.2 +0.5 vs 0.0 2007: -1.0/-0.8 -1.1 +0.1 vs +0.3 ----------------------------------- 2006: -1.0/+0.4 +0.3 -1.3 vs +0.1 2005: +1.0/+0.4 -0.1 +1.1 vs +0.5 2004: +0.7/+0.5 +0.6 +0.1 vs -0.1 2003: -0.9/-0.4 +0.2 -1.1 vs -0.6 2002: +1.0/+0.7 +0.9 +0.1 vs -0.2 ------------------------------------- Analysis: 1) 2013-22: - BoM had no net bias (average miss of 0.0) with 3 significantly too warm and 3 significantly too cold - Euro had a strong warm bias (average miss of +0.6) with 6 of 10 significantly too warm - BoM cooler than Euro 10 of 10 and significantly colder 8 of 10 years 2) 2007-12: - BoM had a strong warm bias (average miss of +0.8) with 5 of 6 significantly too warm - Euro had a moderate warm bias (average miss of +0.35) with 2 of 6 significantly too warm - BoM warmer than Euro 5 of 6 3) 2002-06: based on JAS, not ASO - BoM had a moderate cold bias (average miss of -0.4) with 2 of 5 significantly too cold and 1 of 5 significantly too warm - Euro had no net bias (avg miss ~0) with 1 of 5 significantly too cold and 1 of 5 significantly too warm ------------------------------ Conclusions: 1. Based on the last 10 years, which is the most relevant period, the BoM has had no net bias vs the Euro's strong warm bias, which is what I earlier thought. The Euro hasn't been too cold since way back in 2004! 2. However, the BoM lack of a net bias doesn't mean it hasn't had too warm or too cold years. 3. May of 2023 is highly unusual in that the May BoM for ASO is way up at +2.5. This is a whopping 0.9 warmer than any other back to 2002! The second warmest was the +1.6 of 2015. That turned out to be 0.6 too cold. May of 2023's BoM's +2.5 for ASO is highly unusual also because it isn't just warmer than the Euro, it is MUCH warmer than the Euro's +1.8! The BoM hadn't been warmer than the Euro since way back in 2012 and this time it is 0.7 warmer! There's the chance that BoM had a major modification since 2021 that warmed it. 4. Keeping in mind points #2 and #3 above, considering that the May BoM predicted June to be +1.25 and seeing that it likely will start June no warmer than +0.5, and considering history, the odds are very high that BoM will verify too warm and possibly significantly too warm in June. This all tells me that BoM's +2.5 for ASO will likely verify significantly too warm. Being that the May Euro hasn't been too cold since 2004, I'm predicting 2023 won't be either. Thus, I'm predicting that ASO will end up no warmer than +1.8 and quite possibly cooler due to strong Euro warm bias.

Today's SOI is -21. It will rise sharply again tomorrow. The -SOI streak may or may not end tomorrow as it looks to be close. If not, it looks to end Sunday. It will fall back at the end of the month and is progged to drop to ~~-22 on 5/31. May is projected to end up ~-14 to -15. Looking further, I see no significant AN SLP in Darwin on the models during 6/1-7. It is looking to average NN to possibly a bit BN. Tahiti is less clear as all show BN early while some (operationals) suggest mainly BN probably lingering the whole week while ensembles suggest mainly NN returning. So, whereas the consensus suggests no return to another lengthy strong -SOI the first week of June, it is hard to predict about where that period will end up. For now, I'm leaning fairly neutral.

The one that I'm least impressed by is the MH prediction of 1-4. Since 1950, 55 of 73 seasons (75%) have had 1-4 MH. I mean if I were to pick a number from 1 to 100, asked someone to guess it, allowed for 75 guesses, and one of their 75 guesses turned out to be the number I picked, I wouldn't exactly be impressed about them guessing the number! That's equivalent to their MH outlook.

NOAA issued their "outlook" today with wide ranges: 12-17 storms, 5-9 H, and 1-4 MH. Based on 1950+ stats, NOAA is not making a forecast but them calling it an "outlook" is fair game I suppose. - Since 1950, 33 of 73 (45%) years have had 12-17 storms. - Since 1950, 43 of 73 (59%) have had 5-9 H. - Since 1950, 55 of 73 (75%) have had 1-4 MH.

The forecast for KSAV is for a 24 hour high in the middle 60s on Saturday. That would be a record low high just below the record of 68 set in 1972, another incoming El Niño year.

Refreshing dewpoints mainly in the upper 40s/50s have dominated since yesterday morning and are expected to continue through the weekend. This long period of dewpoints this low is rare for this area in late May. The last time something even close to this was in 2013.

Although I agree with you that the BoM overall hasn't been strongly warm biased when averaging all recent years unlike several other models such as the Euro, being off 0.6 that year is pretty significant and not something to minimize. I tend to think of errors in forecasts issued in spring of 1/2 degree+ by any model in any one year as significant. We have still another very strong -SOI today with -45.

Whereas the past years' verifications support snowman19's assertion of a lack of significant warm bias of the BoM, especially compared to many of the other models, I assume you agree that that doesn't mean never too warm but rather a neutral bias when averaged out. So, sometimes too cold, sometimes or even often about right, and sometimes too warm. Consistent with this, I think you're going to end up correct regarding BoM being too warm for June of 2023 in its May forecast. With 3.4 still only near +0.5 as of last week and with no indication of a rise from there yet, it would appear that 3.4 will start June near +0.5. Based on extrapolating from there, a +1.25 for June as a whole would require much of the mid to late month to be +1.5+. Even with this SOI plunge and considering your thoughts about the current setup not being consistent with past sub -9 SOI Mays, the odds of 3.4 anomaly warming up a full degree C within two weeks are very low based on the past. For any oncoming El Niño, the fastest anomaly warming within two weeks that I could find was right at 1.0, which was from 3/5/2014 to 3/19/2014. The subsequent week warmed another 0.4 and the week after warmed still another 0.1. So, it warmed 1.5 in just four weeks. That's the kind of warming needed for the May BoM forecast for June to verify and that was a very rare event. Moreover, the starting point of that rise was down at -1.2.

I was wrong about yesterday being the low of the stretch. Today's -65 was even lower than yesterday's -62! Today's -65 is the 7th lowest daily SOI on record (back to June of 1991). The only ones lower were five days in 1997-8 and one day in 2009-10 (2/5/2010). And keep in mind it is only May 24th, meaning the current Nino is just starting! **Edit: Though nothing like the rarity now occurring, it now looks like there will be another strong -SOI dip this month with it likely centered on 5/31 near -34.

As mentioned, today's well predicted very low SOI is the 12th lowest since dailies started in June of 1991. The seasons since 1991-2 with the 11 lower SOI dailies were 2009-10, 2006-7, 2004-5, 1997-8, and 1991-2. Of these five seasons, one was super strong Nino, two were strong Nino, and two were weak Nino. So, they were all over the board in terms of El Niño strength with ONI peak ranging from +0.7 to +2.4.

Yesterday's was -43 and today's is -62, which is the lowest since 2/6/10 and 12th lowest since dailies started in June of 1991. Though that will end up the low of this stretch, the next two days will still be strong -SOIs (likely sub -25).

Yep, I remember that like it was yesterday! I'm sure you remember Ken (RIP) closing with "We are the Weather Warriors".

Based on past instances following a sustained strong SOI drop like we've had and which will last a couple more days before rising substantially making it a 17 day solid -SOI, the weekly 3.4 SST anomalies would often rise a few tenths ~3-4 weeks afterward. In this case, that would mean early to mid June. As of last week, 3.4 was still at +0.5. Let's see whether or not there's a rise to +0.8 or so by mid June. Keeping in mind what you said about this case differing quite a bit from other sub -9 Mays, I realize that the typical reaction in Nino 3.4 may not occur. It will be interesting to follow!

The volatile weekly Nino 1+2 region SST anomaly just released (for last week) dropped pretty sharply from +2.4 to +1.7, the coolest in two months. The other three regions changed little or none. The daily SOI plunged to -43. It will be even lower tomorrow, which should be the low point of the month.

But today he posted like the traditional JB as a result of a cool 6Z CFSv2 run for DJF resulting largely from a strong -NAO:

Regardless, the correlation between a sub -9 May SOI and whether or not an oncoming Nino actually occurs has been pretty strong. Since 1950, there have been 14 sub -9 Mays. The only ones not followed by El Niño were the -11.7 of 2005 (that was attributed to lingering effects of the 2004-5 Nino) and the -9.9 of 2001. So, 2001 is the only one of the 14 without explanation. Is there a website to follow the daily EOI? Also, I'd like to see historical EOI data to analyze that is comparable to the SOI historical data that goes way back into the mid 1800s. Are there specific locations' SLPs that determine the EOI like Tahiti and Darwin of the SOI?

Though far from perfect, I've found based on looking at many years of SOI/SST anomaly data that monthly SOI drops are often a pretty good 30-60 day leading indicator for Nino 3.4 SST anomaly rises. So, with a solidly -SOI in May, I'll be looking to see if there's a significant 3.4 anomaly rise in June-July.

SOI monthlies in 2014: After an early solid -SOI in Mar and an accompanying notable Nino 3.4 SST anomaly rise in early spring, Apr-Jul SOIs were ~neutral and SST anomalies became steady. Then -SOI finally came back solidly Aug-Dec, which coincided well with SST anomalies finally rising Sep onward: Jan +11 Feb -2 Mar -12 Apr +6 May +4 Jun -1 Jul -4 Aug -10 Sep -7 Oct-Dec -8 2015: -SOIs dominated entire year with 3 month largest dip (Aug-Oct) way down at -19. SOI didn't rise to neutral til May of 2016, when SST anomalies finally rose to neutral.

Today's SOI came in at ~-26. I'm very roughly estimating the next four days to come in at -35, -57 (lowest daily since 2/6/2010), -50, and -33 followed by a rapid rise. We appear to be looking at ~-12 for May as a whole, a solid indicator for an upcoming Nino of difficult to predict magnitude.

Based on the site noted below, the last two days have been the coldest on record for the date (back to 1958) in the Arctic (80-90N) at 4-5 C BN, which is just colder than 2015, 1962, and 1964. https://ocean.dmi.dk/arctic/meant80n.uk.php The AO has been way up at +2.6 to +2.7 the last two days, the highest on record for the period (since 1950).

On May 10, @bluewaveposted this: "I know that these seasonal forecasts aren’t great. But they are forecasting the -PDO to weaken a bit in the coming months. This would be expected if some of the stronger forecasts for the El Niño worked out." I responded to his post with this: "That's quite a rise forecasted for the PDO between April and July with still another 5 months to go even after that before winter starts. That suggests that a rise to low -1s is quite possible in July. That wouldn't be all that far from +0.50 being that there'd be five months to go." So, being that the recent PDO rise was forecasted, this is following model forecasts thus far. I'm keeping these three past cases of very sharp PDO rises during a new Nino in mind, which tell me there's a small chance that the upcoming DJF PDO will exceed +0.5: 1. Nov-Apr 1883-4 was -1.4. DJF of the very cold 1884-5 was way up at +1.98. That's a 3.38 rise! 2. Nov-Apr of 1975-6 was -1.9. DJF of 1976-7 was up at +1.32. That's a rise of 3.22! 3. Nov-Apr of 2001-2 was -1.3. DJF of 2002-3 was up at +1.42 for a rise of 2.72!

Interesting! Could this be indirectly related to the cold bias of most models in the E US, especially in winter, that results from underestimation of the southeast ridge that has been cited by some pro mets as being largely due to the underestimation of tropical forcing from the near record warm Maritime Continent/W Pacific? I've learned about the idea that record warmth in the MC often makes the atmosphere in the US act as if the MJO is in the MC phases even when it isn't officially in addition to lengthening and strengthening actual MC phases. So with MC MJO phases tending to favor an SER, the models underestimating the forcing from the MC means underestimating the SER. This appears to be the main reason why it is underestimated though the underestimate of the blocking from the near record +AMO is likely an additional factor. Between these two factors, the SER is strongly underestimated, which often results in a NW shift in storm tracks and other features as model forecast time gets closer and closer. So if anyone ever wondered why modeled storm tracks so often shift NW, this appears to be the explanation. So, back to my original question to Bluewave, @jconsorand others. Is what they're talking about here related at all to the same underestimate of MC/W Pac forcing from record warmth that causes the SER to be undermodeled?

Thanks. Per ONI I still favor a strong fall/winter peak but with high end moderate (+1.3 to +1.4) on the low side and super (+2.0+) on the high side. That puts me +1.6 to +1.9 for most likely. But per RONI assuming a -0.4 adjustment due to very warm overall tropical SSTAs, anywhere from a high end weak (+0.9) to low end super (near +2.0) is reasonably possible for the fall/winter peak with +1.2 to +1.5 for most likely. ------------ Redifining Nino indices in a warming climate (relates to RONI): https://iopscience.iop.org/article/10.1088/1748-9326/abe9ed/pdf