donsutherland1

I'm not arguing and have never argued that site changes didn't play any role in the shift in Denver (initial comment about Denver and afterward). I cited Denver, as it's Colorado's largest city, but the frequency in hot days has increased elsewhere, too. My subsequent comments highlighted that far more than site changes are involved in the big recent increase in 90° and 95° days.

Today, Islip recorded its third consecutive high of 74° and Central Park saw its third consecutive high of 73°. The last time both locations had three consecutive days with the same high simultaneously was during July 6-8, 2000 when both locations had three consecutive highs of 79°. Tomorrow will be briefly warmer with highs in the upper 70s to perhaps lower 80s. A few locations could see a shower or thundershower tomorrow night into Friday as a cold front pushes through. The next round of showers or rainfall could be Monday. September 1-15 is on track to achieve a solid cool anomaly. Since 2000, there have been nine years that saw a cooler than normal first half of September. Two-thirds of those years went on to record a warmer than normal second half of September. Only two of those years (2017 and 2019) wound up with a monthly mean temperature of 70.0° or above. Overall, for two-thirds of those years, the cool start was sufficiently cool to produce a cooler than normal monthly anomaly. The last year that saw both a cooler than normal first half and second half of September was 2009. Prior to that, it was 2001. The last year to record a cooler than normal first half of September followed by a warmer than normal September was 2024. The ENSO Region 1+2 anomaly was -0.3°C and the Region 3.4 anomaly was -0.4°C for the week centered around August 27. For the past six weeks, the ENSO Region 1+2 anomaly has averaged +0.33°C and the ENSO Region 3.4 anomaly has averaged -0.32°C. La Niña conditions will likely develop during mid- or late-autumn. The SOI was -5.70 today. The preliminary Arctic Oscillation (AO) was +0.415 today. Based on sensitivity analysis applied to the latest guidance, there is an implied near 59% probability that New York City will have a cooler than normal September (1991-2020 normal). September will likely finish with a mean temperature near 68.4° (0.8° below normal). Supplemental Information: The projected mean would be 0.4° above the 1981-2010 normal monthly value.

My points are as follows: 1) Denver has seen a dramatic increase in hot days. I used 90° highs as an example. The increase is more than just a result of a location change, even as location changes have an impact, otherwise there wouldn't be warming trends across the entire state. Indeed, I've often noted that adjusted data is superior to raw data, precisely because sites move, environments change, times of observation may change, etc. for climate purposes. 2) The broad idea of a greater frequency of hot days does not negate the existence of localized climates or microclimates. Moreover, that idea applies beyond Denver. For example, Boulder has seen half of years with 50 or more 90° or above days and four of its five years with 20 or more 95° or above highs occur since 2000. For Pueblo, 14 of its 17 years with 80 or more 90° days and 9 of its 10 years with 50 or more 95° days have occurred since 2000. In short, in Colorado's populated areas, the 2000s have seen a disproportionate share of hot days. No other period rivals the 2000s in terms of the frequency of such heat. 3. The old Stapleton Airport does not have a continuous record. The old site was demolished and replaced with Central Park. A new ASOS was installed around 2022 there.

Staptleton Airport was decommissioned, demolished, and replaced by Denver Center Park. The environment is quite different from what it was when record-keeping ended in 1995 at Stapleton Airport. The environment for the new ASOS isn't comparable to what it was at Stapleton Airport. The warming summers since 1995 isn't an artifact of Denver International Airport. It is a statewide phenomenon. 12/15 (80%) of Colorado's summers with a mean temperature of 67.0° or above and all nine of its hottest summers have occurred since 2000. Five of the last six summers (2023 being the exception) have had a mean temperature of 67.0° or above. Statewide maximum temperatures have risen somewhat faster than statewide mean temperatures.

Tomorrow will be variably cloudy with a few showers possible in places. Highs will top out mainly in the upper 60s. Thursday could be briefly warmer with highs in the upper 70s to perhaps lower 80s. A scattered shower or thundershower is possible Thursday night into Friday as a cold front pushes through. September 1-15 is on track to achieve a solid cool anomaly. Since 2000, there have been nine years that saw a cooler than normal first half of September. Two-thirds of those years went on to record a warmer than normal second half of September. Only two of those years (2017 and 2019) wound up with a monthly mean temperature of 70.0° or above. Overall, for two-thirds of those years, the cool start was sufficiently cool to produce a cooler than normal monthly anomaly. The last year that saw both a cooler than normal first half and second half of September was 2009. Prior to that, it was 2001. The last year to record a cooler than normal first half of September followed by a warmer than normal September was 2024. The ENSO Region 1+2 anomaly was -0.3°C and the Region 3.4 anomaly was -0.4°C for the week centered around August 27. For the past six weeks, the ENSO Region 1+2 anomaly has averaged +0.33°C and the ENSO Region 3.4 anomaly has averaged -0.32°C. La Niña conditions will likely develop during mid- or late-autumn. The SOI was -13.78 today. The preliminary Arctic Oscillation (AO) was +0.461 today. Based on sensitivity analysis applied to the latest guidance, there is an implied near 59% probability that New York City will have a cooler than normal September (1991-2020 normal). September will likely finish with a mean temperature near 68.3° (0.9° below normal). Supplemental Information: The projected mean would be 0.3° above the 1981-2010 normal monthly value.

Synoptic scale events can still lead to snowy outcomes even with hostile boundary conditions. The super El Niño winter of 2015-16 is an example. There was a single massive snowstorm that skewed the numbers. There was also a severe Arctic outbreak that sent the mercury in Central Park to -1° on February 14. Much of the rest of the winter was very warm with a lack of snowfall.

Assuming that the SSTs will drive the pattern, he would need big changes from what is forecast for Winter 2025-26 to be similar to that for Winter 2013-14. The core of the warm anomalies would need to shift much farther east. 2013-14 Current Forecast (ECMWF): Current Forecast (CFSv2):

White Plains recorded a low of 53° this morning, its second consecutive day with a low of 53° or below. 2025 has now seen two stretches of at least two consecutive lows of 53° or below through September 9th (August 30-31 and September 8-9). The last time that happened was in 2017 (September 1-2 and September 8-9). Since August 1, White Plains has had 20 days with lows of 59° or below. That is the most since 1994 and tied with 1962 and 1987 for the fourth most. Most Days with Lows < 60° (August 1-September 9): 1. 22, 1964 and 1994 3. 21, 1957 4. 20, 1962, 1987, 2025 7. 19, 1982 8. 18, 1986 9. 17, 1972, 1988, 1997

Much of the region saw its chilliest morning so far this season. Low temperatures included: Binghamton: 44° Bridgeport: 50° Caldwell: 49° Danbury: 44° Farmingdale: 55° Hartford: 45° Islip: 58° Montgomery: 43° New Haven: 50° New York City-Central Park: 56° New York City-JFK Airport: 56° New York City-LaGuardia Airport: 60° Newark: 55° Philadelphia: 56° Poughkeepsie: 44° Trenton: 51° White Plains: 49° (earliest season first 40s since 2001) Temperatures will mainly top out in the 70s during the daytime and fall into the 50s at nighttime in New York City through at least Wednesday. Thursday could be briefly warmer with highs in the upper 70s to perhaps lower 80s. A shower or thundershower is possible on Wednesday as warm front moves across the region and Thursday as a cold front pushes through. The ENSO Region 1+2 anomaly was -0.3°C and the Region 3.4 anomaly was -0.4°C for the week centered around August 27. For the past six weeks, the ENSO Region 1+2 anomaly has averaged +0.33°C and the ENSO Region 3.4 anomaly has averaged -0.32°C. La Niña conditions will likely develop during mid- or late-autumn. The SOI was +0.06 today. The preliminary Arctic Oscillation (AO) was +0.073 today. Based on sensitivity analysis applied to the latest guidance, there is an implied near 61% probability that New York City will have a cooler than normal September (1991-2020 normal). September will likely finish with a mean temperature near 68.1° (1.1° below normal). Supplemental Information: The projected mean would be 0.1° above the 1981-2010 normal monthly value.

No. I checked.

The temperature dipped to 49° at White Plains this morning. Last year's first low in the 40s occurred on September 9. That is the earliest season low in the 40s since 2001 when the temperature fell to 49° on July 3.

FYI, the NWS has added the rainfall amounts that were missing for both August and yesterday for Bridgeport. The region benefits from its high quality and responsive NWS office.

As of 4 pm, Central Park had a high of 68°. Should Central Park have a high of 69° or below, it would be the first time New York City has seen a high in the 60s during the first week of September since September 6, 2019 when the mercury topped out at 67°. Yesterday, Central Park received 1.01" of rain, its first 1.00" or above daily rainfall since July 14. Additional rain fell today. Two-day rainfall amounts across the region included: Islip: 0.77" New Haven: 1.39" New York City-Central Park: 1.64" New York City-JFK Airport: 0.99" New York City-LaGuardia Airport: 1.56" Newark: 0.89" White Plains: 1.56" Temperatures will mainly top out in the 70s during the daytime and fall into the 50s at nighttime in New York City through at least Wednesday. Thursday could be briefly warmer with highs in the upper 70s to perhaps lower 80s. Additional showers or rain is possible on Thursday, as another cold front crosses the region. The ENSO Region 1+2 anomaly was -0.3°C and the Region 3.4 anomaly was -0.4°C for the week centered around August 27. For the past six weeks, the ENSO Region 1+2 anomaly has averaged +0.33°C and the ENSO Region 3.4 anomaly has averaged -0.32°C. La Niña conditions will likely develop during mid- or late-autumn. The SOI was +6.95 today. The preliminary Arctic Oscillation (AO) was -0.932 today. Based on sensitivity analysis applied to the latest guidance, there is an implied near 59% probability that New York City will have a cooler than normal September (1991-2020 normal). September will likely finish with a mean temperature near 68.2° (1.0° below normal). Supplemental Information: The projected mean would be 0.2° above the 1981-2010 normal monthly value.

Years with one or more 90° temperatures on or after the fall equinox for select New York City Area locations: Bridgeport: 1970 Islip: 1980 New York City-Central Park: 1881, 1895, 1914, 1927, 1933, 1938, 1939, 1941, 1959, 1961, 1970, 1980, 2017, 2019 New York City-JFK Airport: 1970, 1980, 2007, 2017, 2019 New York City-LaGuardia Airport: 1941, 1946, 1970, 1980, 1998, 2007, 2010, 2017, 2019 Newark: 1881, 1895, 1908, 1914, 1938, 1939, 1941, 1949, 1958, 1959, 1961, 1968, 1970, 1980, 1998, 2007, 2010, 2016, 2017, 2019 White Plains: 1970, 2017, 2019

It should also be noted that since 2000, there has been a tendency for Newark's above normal days to be backloaded in September (relative to the 1991-2020 baseline): That doesn't mean that this year will be the case, but the idea that there could be much warmer air in the Plains States/Midwest toward mid-month suggests that it's something to watch for should the guidance continue to show such warmth in the Plains States/Midwest. The Week 3 outlooks on the ECMWF Weeklies and the CFSv2 are warmer than normal in the Northeast:

For the second time in less than a month, Bridgeport is not reporting its precipitation. Yesterday's hourly data: The lowered visibilities are consistent with falling rain: Yesterday's NOWData: In sum,Bridgeport had measurable rainfall yesterday. That data was not recorded. So far, today's rainfall has also not been measured.

What this means is that the social media hurricane hypesters will simply shift the goalposts to a few weeks down the road to maybe late September or October (an art that many have perfected from punting long-range cold/snow calls when they don't verify). I suspect that there will be some activity before the season concludes, so they will also proclaim that they were correct. Also, there's no reason the current season can't end with only a modest uptick in activity. Every season does not end with a spectacular burst of late-season storms. There's a window currently suggested on the MJO guidance for the second half of September (probably after the 20th) into the first week of October. Afterward, things could largely shut down. The forecast lift isn't as great or widespread as one would look for if there were to be a big explosion of tropical cyclone genesis or a large number of hurricanes.

Today saw the temperature top out at 90° at Newark and strong thunderstorms bring drenching rainfall to parts of the region. Newark reached 90° or above in June, July, August, and September for the first time since 2023. 2023 also saw the mercury hit 90° in May, unlike this year. Hourly rainfall amounts included: Central Park: 0.89", LaGuardia Airport: 0.78"; and, White Plains: 0.90". Today is Central Park's wettest day since 2.64" fell on July 14. A cold front crossing the region will bring additional showers and thundershowers tonight. Rain could linger into tomorrow as the front slows. A general 0.50"-1.50" rainfall is likely around the New York City area and its nearby suburbs with localized amounts in excess of 2.00" by the time the rain ends tomorrow. Following the passage of the cold front, highs will reach the upper 60s to lower 70s tomorrow. Afterward, temperatures will mainly top out in the 70s during the daytime and fall into the 50s at nighttime in New York City through at least the middle of next week. Additional showers or rain is possible around midweek. The ENSO Region 1+2 anomaly was -0.3°C and the Region 3.4 anomaly was -0.4°C for the week centered around August 27. For the past six weeks, the ENSO Region 1+2 anomaly has averaged +0.33°C and the ENSO Region 3.4 anomaly has averaged -0.32°C. La Niña conditions will likely develop during mid- or late-autumn. The SOI was +17.47 today. The preliminary Arctic Oscillation (AO) was -0.837 today.

0.78" hourly amount at Central Park as of 3:32 pm.

For reference: Highest hourly amount for September 6: 0.96", 2008.

Newark has reached 90° for the 39th time this year. Thunder is also reported.

I suspect that its seeming inability to see cold anomalies except at shorter timescales has a lot to do with the idea that boundary conditions drive seasonal averages. ENSO, PDO are prominent conditions. The oceans overall are warming. Therefore, the model forecasts are tipped toward the warmer idea more broadly than is realistic. Worse, the coefficients of determination for such variables related to boundary conditions and actual seasonal outcomes are very low. These weak relationships reveal that other important factors are involved, including synoptic scale events that cannot be reliably forecast beyond 10-14 days. Some of these additional variables may not yet be known. Synoptic scale events i.e., large snowstorms, Arctic blasts, etc., can have a great influence on the overall seasonal outcomes. Thus, even a warm winter can be much snowier than normal or a cold winter can lack snowfall. On account of these other variables, every La Niña or El Niño event is not alike. The seasonal models are not yet at a stage where they can even begin to consistently resolve the actual events that ultimately produce the seasonal outcome. A similar situation applies to subseasonal forecasting. Not surprisingly, beyond two weeks, model skill on the weekly guidance largely disappears. There also seems to be a larger deal of persistence in the two week or longer forecasts than what actually occurs. AI may improve some of these outcomes. But even then, big challenges could still persist. For example, even as some experiments with random forest models have shown a degree of improved skill in forecasting ENSO, those models are constrained by their knowledge base. Hence, when it comes to forecasting extreme events e.g., super El Niño events, they have great difficulty. Perhaps the combination of AI and quantum computing might produce some significant breakthroughs. But that's still in the future and perhaps a decade or more away, assuming society values science and basic research to make the investments necessary to arrive at that improved state of forecasting. That's an open question in some areas and it will become even more relevant as major states grapple with the costs of aging populations, rising debt relative to GDP, etc., and the trade-offs involved in making budget allocations.

With ECMWF seasonal maps coming out, focus briefly turns to the coming winter. While it is far too soon to expect much skill from the models, recent winter outcomes suggest that New York City is beginning to put distance between the present and a golden snowfall era. The core of winter (December-February) was especially snowy for the most part from Winter 1999-00 through Winter 2017-18. Storms dumped an average of 2.9" per event and there were 30 days with 6" or more snow, including 9 with 10" or more. The December-February periods saw an average of 28.2" of snow. For all other winters, storms averaged 1.9" per event with the December-February period averaging 20.8".

The 0.48" rainfall at Bridgeport today broke the daily record for September 5. The old mark of 0.44" was set in 2012.

Here you go: