Continuing data issues limit a complete discussion of the current global weather-climate situation. I will only be able to speculate about components involving the earth-atmosphere AAM budget (torques, transports, etc.) as well as the GWO.
Global tropical SSTs have been relatively “steady state” since my posting nearly 10 days ago. Very warm ocean waters (high heat content) extend from north of Australia into the South Pacific, having anomaly magnitudes in excess of 2C and totals greater than 30C especially ~10S. The latter have been shifting slowly south with the seasonal/annual cycle. SST anomalies remain well below normal in all Nino regions (~160E-South America), with magnitudes less than minus 3C extending to depths ~100m east of 140W. The weak oceanic Kelvin wave continues to propagate east at around 150m depth per latest 5-day averaged TAO buoy data having reached about 150W. Finally, monitoring suggests the convectively suppressed Indian Ocean is slowly warming while much of the tropical and North Atlantic Ocean remains warmer than climatology.
So what is the point to giving a quick overview of the global tropical SSTs, at least for this posting (see http://www.pmel.noaa.gov/tao/jsdisplay and other relevant links for details)? Although I can just offer speculation here not having objective data sets, the on-going circulation response cannot be attributed to El-Viejo alone. In fact, this is often (always?) the case where the ENSO variability is only a component. Simple monitoring of the global circulation for the last several years in itself demonstrates that Eastern Hemisphere/west Pacific Ocean warm pool SSTAs have had significant impacts onto the global circulation (including right now).
Strong-severe tropical convection currently extends from the central equatorial Indian Ocean east-southeast to just north of Australia into a southwestward shifted South Pacific Convergence Zone (SPCZ). The core of this tropical convective forcing is ~10S/130E, also shifting south with the seasonal cycle. Suppression exists across the region of the equatorial date line. OLR anomaly magnitudes are in excess of 50 W/m**2.
There is a significant MJO component to this tropical forcing, located in phase 5 per WH(2004) phase space plots updated through 29 December. However, as suggested in my last posting, atmosphere-ocean coupling involving the very warm SSTs north of Australia appears to have been in progress for at least the last week. Not only has the WH(2004) MJO signal slowed down, other plots such as Hovmollers of OLR/A and velocity potential suggest a recent stationary signal. However, I do think a relatively fast MJO dynamical signal will emerge into the Western Hemisphere during the next few weeks. Again, there is HUGE uncertainty about timing and other details.
Animations of upper tropospheric daily mean vector wind anomalies show the classic expected response given the spatial distribution of the tropical forcing discussed above. Twin subtropical anticyclones (cyclones) are centered ~110E (~date line) with anomaly magnitudes in excess of 40m/s for the equatorial Pacific westerlies. There have been several episodes of meridionally oriented Rossby wave energy dispersions (RWDs) into the extratropics tied to circulation features such as these for the last several weeks. For the PNA sector these RWDs have led to robust cyclonic circulation anomalies around Alaska poleward of a central Pacific Ocean anticyclone. Several strong troughs have impacted the USA west coast on into the Plains as a result, accompanied by high impact weather events including ice and snowstorms.
For the Asian-North American extratropics, this response fits ~phases 3-4 for the WB(200?) DJF MJO composite (a component of the GWO; publication is planned), but it is somewhat different for the tropics. The point is there are highly complex non-linear dynamical feedback processes involving the extratropics and particularly the synoptic eddies going on. For instance, during the last ~5 days a fast baroclinic wave packet tied to presumably a positive East Asian mountain torque (explained by the GWO) has been aimed at North America. This will lead to the west coast ridge amplification and anomalous cold for the central/eastern USA the models have been predicting. However, this is a transient event and when updated, all that is probable to be observed is at least one circuit around GWO phase 3 (old GSDM Stage 1).
The East Asian jet extension as a response to the positive mountain torque will strongly impact the USA west coast by the end of this upcoming week per all models. As shown by DJF GWO and MJO composites WB are working on, the MJO itself does not generate enough extratropical westerly wind flow to allow the East Asian jet to impact the USA west coast. A strong positive East Asian mountain torque needs to be involved, on average.
Uncertainty is huge where the global circulation goes from here, especially when lack of resources denies me much needed data sets (latter speaking for myself). I think GWO phase 3 best describes the current global weather-climate situation. As long as the tropical convective forcing stays coupled north of Australia, I do not think this subseasonal base state (La-Nina understood) will change. However, these coupling events have recently lasted ~2 weeks, and several tropical cyclones are developing/occurring with the convection. As mentioned above, I do think a MJO component will continue into the Western Hemisphere during the next several weeks. Support for this notion are an intense westerly wind burst (WWB) north of Australia heading into the South Pacific, as well as the strong upper tropospheric equatorial westerly wind flow anomalies also mentioned above. Whether or not this WWB will be close enough to the equator (may be too far south) to generate another downwelling oceanic Kelvin wave is unclear.
Similar to about the first half of December 2007, I suspect any larger GWO orbit through phases 5-8 that may occur ~mid/late January 2008 will be relatively fast. Hence, for the lower 48 states, weeks 2-3 may have similar weather to that observed during much of December. Delivery of Arctic airmasses will be limited as long as large cyclonic circulation anomalies persist near Alaska. Nevertheless, particularly given January climatology, extreme winter weather may occur from the west coast-central/southern Rockies into the Plains. Heavy rainfall and even severe local storms may be a concern for especially the lower Mississippi and Ohio Valleys. Should phases 5-6 of the GWO occur, say by ~ week 3, Arctic air may spill into the USA. Predicting the latter involves daily monitoring; accept it!!! Finally, on more of a seasonal time scale, “bucking” the La-Nina composite signal, my subseasonal monitoring tells me not to be so pessimistic in regard to dryness across the southwestern states into the Central/Southern Plains for JFM. In fact, precipitation anomalies for portions of California through Desert Southwest and Central/Southern Plains may end up being positive for JFM if on-going feedback processes do not change.
Internationally, heavy-severe rainfall/thunderstorms will be a concern for Eastern Hemisphere areas already discussed above for weeks 1-2. These impacts may spread into the South Pacific Islands and eventually tropical South America ~ weeks 2-4 (could be sooner). Yes, the sprit of “Gabrielle” is currently festering west of Africa.
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH(2004) “convention”. In addition, the GWO plot does not have the ENSO signal removed.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will attempt another discussion next weekend.
Ed Berry
Sunday, December 30, 2007
Friday, December 21, 2007
Typical El-Viejo??? - Update
There has been no significant change to the distribution of global tropical SSTs since my 15 December post, particularly across the Indo-Pacific region. All Nino regions have well below normal SSTs ~minus 1-3C extending to at least 100-150m deep, flanked by the warm Pacific Ocean horseshoe. An oceanic Kelvin wave with anomalies roughly plus 1-3C continues to propagate east (~2.5 m/s) through the region of 180-160W at around 150m depth per 5-day averaged TAO buoy data. While much of the equatorial and southern Indian Ocean has cooled to less than 28C (anomalies ~minus 1-2C) due to recent moist convection, the very warm waters from north of Australia to the South Pacific Ocean remain. Totals in excess of 30C are common across the latter region.
The Eastern Hemisphere tropical convective forcing has become somewhat better organized during the last few days. Full disk satellite imagery shows this large region of intense anomalous and at times severe tropical rainfall extending from the Arabian Sea to a concentrated area over central Indonesia (~0/120E) into the South Pacific Ocean. Various monitoring tools indicate there have been several components contributing to this tropical forcing. However, as shown by the WH2004 phase space and coherent modes Hovmoller plots, there is a significant MJO contribution (currently MJO phase 3-4). My own back of the envelop phase speed calculation has this convective region shifting east at ~4-5m/s since early this month. I am expecting a roughly 1-2 week period of coupling to the very warm 120E-180 SSTs by early January, before a MJO signal in all likelihood propagates into the Western Hemisphere. Again, I remind the readers there is always LARGE uncertainty about timing.
The global circulation is very much La-Nina like, which can be thought of as an enhanced climatological base state. Global relative AAM is roughly 2 standard deviations below the R1 data climatology (through 18 December) having well above average zonal mean easterly wind flow anomalies across the equatorial and subtropical atmospheres. The GWO is essentially orbiting around phase 3 (old GSDM Stage 1). However, there are always issues of non-linear dynamical feedbacks involving the extratropical eddies. For instance, per animations of daily mean upper tropospheric vector wind anomalies, while twin subtropical anticyclones (cyclones) have become established ~100-120E (~date line), there has been strong Rossby wave energy dispersions (RWDs) into the global extratropics. These RWDs have been very fast (~30m/s) not only leading to the barrage of western USA troughs, but also feeding blocking structures over Scandinavia and very recently northeast Asia. The latter will lead to a build-up of bitterly cold Arctic across Alaska during the next few weeks.
A strong zonal mean poleward AAM transport signal has become evident around 30N during the last few days, with flux divergence from the subtropics and convergence ~35N. In addition to the eddies, this transport signal has also been adding anomalous westerly flow to the northern extratropics (including the North Pacific Ocean). My point is the appearance of an AAM transport signal such as this adds support (through exchange processes involving the surface torques) to the predicted PNA synoptic evolutions I discussed a week ago.
I think there will be ~couple roughly 10-day circuits around GWO phase 3 until early-mid January when a large push to GWO phase 5 (GSDM Stage 2) may occur. That would be most probable after the period of atmosphere-ocean coupling discussed above. An evolution through GWO phases 6-8 (GSDM Stages 3-4) is possible afterwards (~weeks 4-6). Should the latter occur, more significant precipitation for locations such as California and the Desert Southwest may occur.
Hence there is no change to the predictive insights I offered in my 15 December discussion. Several troughs riding a strong North Pacific jet stream (yes, this can happen during a La-Nina) will continue to impact most of the country through early January 2008. I do think it is probable to see large ridge amplification from the east Pacific/off the USA west coast into Alaska by ~mid January 2008. Interestingly, the ESRL/PSD week-2 ensemble mean of 500mb height anomalies suggests this possibility while the NCEP GFS ensemble mean does not. However, even though the week-2 models have been “all over the place” recently, I think there will be upcoming better agreement on this GWO phase 5 scenario. If this happens, severe Arctic air may impact the Plains and eventually the eastern USA by ~mid January, after possibly initially spilling into the Pacific Northwest.
Internationally, severe rainfall and thunderstorms are probable to continue across portions of Australia and Indonesia and even into the Southwest Pacific during the next couple weeks. The north and northeast coast of Australia may become vulnerable to tropical cyclones weeks 2-3. Europe should become active as the Scandinavian blocking breaks down while bitterly cold air even by Siberia standards dominates northeast Asia. Again, some of the latter airmass may plunge into the USA via western Canada by ~mid January 2008.
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 “convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will not be able to post another discussion until late next weekend at the soonest.
Ed Berry
The Eastern Hemisphere tropical convective forcing has become somewhat better organized during the last few days. Full disk satellite imagery shows this large region of intense anomalous and at times severe tropical rainfall extending from the Arabian Sea to a concentrated area over central Indonesia (~0/120E) into the South Pacific Ocean. Various monitoring tools indicate there have been several components contributing to this tropical forcing. However, as shown by the WH2004 phase space and coherent modes Hovmoller plots, there is a significant MJO contribution (currently MJO phase 3-4). My own back of the envelop phase speed calculation has this convective region shifting east at ~4-5m/s since early this month. I am expecting a roughly 1-2 week period of coupling to the very warm 120E-180 SSTs by early January, before a MJO signal in all likelihood propagates into the Western Hemisphere. Again, I remind the readers there is always LARGE uncertainty about timing.
The global circulation is very much La-Nina like, which can be thought of as an enhanced climatological base state. Global relative AAM is roughly 2 standard deviations below the R1 data climatology (through 18 December) having well above average zonal mean easterly wind flow anomalies across the equatorial and subtropical atmospheres. The GWO is essentially orbiting around phase 3 (old GSDM Stage 1). However, there are always issues of non-linear dynamical feedbacks involving the extratropical eddies. For instance, per animations of daily mean upper tropospheric vector wind anomalies, while twin subtropical anticyclones (cyclones) have become established ~100-120E (~date line), there has been strong Rossby wave energy dispersions (RWDs) into the global extratropics. These RWDs have been very fast (~30m/s) not only leading to the barrage of western USA troughs, but also feeding blocking structures over Scandinavia and very recently northeast Asia. The latter will lead to a build-up of bitterly cold Arctic across Alaska during the next few weeks.
A strong zonal mean poleward AAM transport signal has become evident around 30N during the last few days, with flux divergence from the subtropics and convergence ~35N. In addition to the eddies, this transport signal has also been adding anomalous westerly flow to the northern extratropics (including the North Pacific Ocean). My point is the appearance of an AAM transport signal such as this adds support (through exchange processes involving the surface torques) to the predicted PNA synoptic evolutions I discussed a week ago.
I think there will be ~couple roughly 10-day circuits around GWO phase 3 until early-mid January when a large push to GWO phase 5 (GSDM Stage 2) may occur. That would be most probable after the period of atmosphere-ocean coupling discussed above. An evolution through GWO phases 6-8 (GSDM Stages 3-4) is possible afterwards (~weeks 4-6). Should the latter occur, more significant precipitation for locations such as California and the Desert Southwest may occur.
Hence there is no change to the predictive insights I offered in my 15 December discussion. Several troughs riding a strong North Pacific jet stream (yes, this can happen during a La-Nina) will continue to impact most of the country through early January 2008. I do think it is probable to see large ridge amplification from the east Pacific/off the USA west coast into Alaska by ~mid January 2008. Interestingly, the ESRL/PSD week-2 ensemble mean of 500mb height anomalies suggests this possibility while the NCEP GFS ensemble mean does not. However, even though the week-2 models have been “all over the place” recently, I think there will be upcoming better agreement on this GWO phase 5 scenario. If this happens, severe Arctic air may impact the Plains and eventually the eastern USA by ~mid January, after possibly initially spilling into the Pacific Northwest.
Internationally, severe rainfall and thunderstorms are probable to continue across portions of Australia and Indonesia and even into the Southwest Pacific during the next couple weeks. The north and northeast coast of Australia may become vulnerable to tropical cyclones weeks 2-3. Europe should become active as the Scandinavian blocking breaks down while bitterly cold air even by Siberia standards dominates northeast Asia. Again, some of the latter airmass may plunge into the USA via western Canada by ~mid January 2008.
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 “convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will not be able to post another discussion until late next weekend at the soonest.
Ed Berry
Saturday, December 15, 2007
AAM Trend Still in Sync with the Financial Markets
I wish the upward trend in the USA financial markets a week ago would have continued this last work week. This makes me wonder if any statistically significant correlations exist between subseasonal-interannual atmospheric variability and the world markets. The former would be in regard to the GWO and MJO while the latter would relate to ENSO.
This posting only updates what I have been writing during the last few weeks. The spatial pattern of tropical SSTs across the Indo-Pacific sector is consistent with a mature La-Nina. Anomalies along the equator from 160E-South America remain negative with magnitudes at least minus 2.5C starting at 140W (per TAO buoy data). At depth these negative anomalies are at least 150-200m deep while comparable positive anomalies (~3C) remain in the region of the equatorial date line. The latter is a response to a deepening of the oceanic thermocline forced by a Kelvin wave initiated by the recent past MJO.
While the equatorial Indian Ocean has cooled during the last week from increased tropical convective rainfall, very warm SSTs are present from about the north coast of Australia to Indonesia into the South Pacific Ocean. SST anomalies are ~plus 1-2C from around 120E-180 south of the equator with totals generally in excess of 30C (86F). These very warm Eastern Hemisphere tropical SSTs will have an important contribution to the evolution of the global circulation during the next several weeks.
Per full disk satellite imagery tropical convective forcing has become very intense across the Indian Ocean centered ~5S/80-90E having OLRA less than minus 70W/m**2 per 3-day averaged BMRC plots. Compensating strong suppression exists downstream from Indonesia-date line. This forcing is tied to the MJO, with latest WH2004 plots projecting onto phase 3 due to RMM2 being strongly negative. For the time being, the MJO signal has stalled. The GWO (which also captures the circulation responses due to the MJO in addition to extratropical non-oscillatory forcing) has remained in-sync with the tropical convective forcing returning to the Indian Ocean. Phase 3 (old GSDM Stage 1) best describes the projection of the current global circulation onto the GWO. Anomalous trades from Indonesia-date line downstream from the convection are reinvigorating La-Nina.
About a week ago global relative AAM tendency crashed to ~minus 50 Hadleys (3-day average/R1 data climatology). As of 13 December this tendency did spike to ~minus 5 Hadleys with a large component coming from a renewed positive global frictional torque. The latter was ~plus 10 Hadleys, seemingly exhibiting a recent periodicity of ~60 days (consistent with the GWO). Strengthening trades throughout the global subtropics have contributed. However, global relative AAM has again become quite low ~minus 2 standard deviations with the redevelopment of intense zonal mean easterly wind flow anomalies (~5m/s at 200mb) throughout the global equatorial and subtropical atmospheres. Also helping to remove westerly flow from the atmosphere are the large north-south mountains, with the global mountain torque ~minus 15 Hadleys having a large component from East Asia.
Animations of upper tropospheric daily mean vector wind anomalies show the expected circulation characteristics per above. Twin subtropical anticyclones are getting better defined ~90-120E with downstream cyclones in the region of the equatorial date line. Westerly wind flow anomalies (~10-20m/s) are developing across the Western Hemisphere particularly across the Pacific Ocean. In fact, some of this upper tropospheric westerly wind flow is being directed toward the southwestern USA.
Zonally oriented rapid Rossby wave energy dispersions (RWDs) directed toward the USA emanating from the Indian Ocean tropical forcing have been a dynamical component to the recent stormy winter weather pattern across the country. Per EMC/NCEP web site visual verification of predicted NCEP/GFS week-2 ensemble means of 500mb height anomalies valid roughly the last 7 days verses observed did not capture the RWDs leading to the recent negative height anomalies across the Desert Southwest. The latter is not only a systematic bias issue, but also the inability of global models to accurately predict tropical convective forcing after about day 5. I speculate that numerical model performance at lead times greater than ~3 days may remain "unusually problematic" for roughly the next 1-3 weeks.
I suspect the GWO may exhibit one or two small roughly 5-10 day circuits around phase 3 (consistent with El-Viejo) before resuming its large orbit toward phase 5 (GSDM Stage 2) and beyond. The WH2004 MJO phase space plot will likely do the same before showing an eastward propagation signal once again. One reason I feel this way is that I think the Indian Ocean tropical convective forcing is likely to shift east-southeast toward the very warm SSTs north of Australia during the next few weeks.
Of course, as always there are large uncertainties about timing and synoptic details. However, I think it is probable for both the GWO and MJO to be in roughly phase 5 by early January 2008, as I discussed a week ago. The seasonal cycle will also have “its say in this matter”, since climatologically SSTs 29C and greater become covered with tropical convection during the first part of January. So, are we going to have a period of ocean-atmosphere coupling ~120E-180 south of the equator in about 3 weeks (with the onset of the Australian monsoon)? Are we going to see another decent GWO/MJO circuit afterwards? The answers to these types of predictability questions are unclear; however, I speculate yes.
A generally active weather regime is probable to continue for most of the country through at least the end of 2007. As I typed a week ago, the period from around Christmas-early January 2008 may be exceptionally stormy focusing on the Rockies and Plains in the presence of an Arctic cold air source. However, both coasts and Alaska are probable to have impacts from our active regime. I also think it is probable to see strong ridge amplification (greater than climatology) from just west of North America-Alaska by mid January per above reasoning. The latter would suggest an anomalous cold regime from east of the Rockies-east coast with the possible exception of the Deep South.
Internationally, the main tropical cyclone hazard should shift from the South Indian Ocean (Bay of Bengal understood) toward the region of Australia and eventually southwest Pacific Ocean weeks 2-3. Surface westerly wind anomalies ~10-15m/s in that region certainly support this notion. By that time the Australian monsoon should be on-going. Indonesia and even the Philippines are also probable to have possibly severe rainfall events by weeks 2-3. I would think the current Scandinavian blocking pattern should break down by week 3, possibly retrograding into the North Atlantic Ocean. In any case, cool/wet conditions seem probable for much of the Mediterranean region through week-2. Arctic air will continue to build up and expand across particularly Siberia as well as Alaska and much of Canada during the next few weeks. Finally, I hope the spirit of Gabrielle will rest for the warm North Atlantic basin until boreal summer 2008.
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 “convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I may not be able to post another discussion for at least a couple weeks; stay tuned.
Ed Berry
This posting only updates what I have been writing during the last few weeks. The spatial pattern of tropical SSTs across the Indo-Pacific sector is consistent with a mature La-Nina. Anomalies along the equator from 160E-South America remain negative with magnitudes at least minus 2.5C starting at 140W (per TAO buoy data). At depth these negative anomalies are at least 150-200m deep while comparable positive anomalies (~3C) remain in the region of the equatorial date line. The latter is a response to a deepening of the oceanic thermocline forced by a Kelvin wave initiated by the recent past MJO.
While the equatorial Indian Ocean has cooled during the last week from increased tropical convective rainfall, very warm SSTs are present from about the north coast of Australia to Indonesia into the South Pacific Ocean. SST anomalies are ~plus 1-2C from around 120E-180 south of the equator with totals generally in excess of 30C (86F). These very warm Eastern Hemisphere tropical SSTs will have an important contribution to the evolution of the global circulation during the next several weeks.
Per full disk satellite imagery tropical convective forcing has become very intense across the Indian Ocean centered ~5S/80-90E having OLRA less than minus 70W/m**2 per 3-day averaged BMRC plots. Compensating strong suppression exists downstream from Indonesia-date line. This forcing is tied to the MJO, with latest WH2004 plots projecting onto phase 3 due to RMM2 being strongly negative. For the time being, the MJO signal has stalled. The GWO (which also captures the circulation responses due to the MJO in addition to extratropical non-oscillatory forcing) has remained in-sync with the tropical convective forcing returning to the Indian Ocean. Phase 3 (old GSDM Stage 1) best describes the projection of the current global circulation onto the GWO. Anomalous trades from Indonesia-date line downstream from the convection are reinvigorating La-Nina.
About a week ago global relative AAM tendency crashed to ~minus 50 Hadleys (3-day average/R1 data climatology). As of 13 December this tendency did spike to ~minus 5 Hadleys with a large component coming from a renewed positive global frictional torque. The latter was ~plus 10 Hadleys, seemingly exhibiting a recent periodicity of ~60 days (consistent with the GWO). Strengthening trades throughout the global subtropics have contributed. However, global relative AAM has again become quite low ~minus 2 standard deviations with the redevelopment of intense zonal mean easterly wind flow anomalies (~5m/s at 200mb) throughout the global equatorial and subtropical atmospheres. Also helping to remove westerly flow from the atmosphere are the large north-south mountains, with the global mountain torque ~minus 15 Hadleys having a large component from East Asia.
Animations of upper tropospheric daily mean vector wind anomalies show the expected circulation characteristics per above. Twin subtropical anticyclones are getting better defined ~90-120E with downstream cyclones in the region of the equatorial date line. Westerly wind flow anomalies (~10-20m/s) are developing across the Western Hemisphere particularly across the Pacific Ocean. In fact, some of this upper tropospheric westerly wind flow is being directed toward the southwestern USA.
Zonally oriented rapid Rossby wave energy dispersions (RWDs) directed toward the USA emanating from the Indian Ocean tropical forcing have been a dynamical component to the recent stormy winter weather pattern across the country. Per EMC/NCEP web site visual verification of predicted NCEP/GFS week-2 ensemble means of 500mb height anomalies valid roughly the last 7 days verses observed did not capture the RWDs leading to the recent negative height anomalies across the Desert Southwest. The latter is not only a systematic bias issue, but also the inability of global models to accurately predict tropical convective forcing after about day 5. I speculate that numerical model performance at lead times greater than ~3 days may remain "unusually problematic" for roughly the next 1-3 weeks.
I suspect the GWO may exhibit one or two small roughly 5-10 day circuits around phase 3 (consistent with El-Viejo) before resuming its large orbit toward phase 5 (GSDM Stage 2) and beyond. The WH2004 MJO phase space plot will likely do the same before showing an eastward propagation signal once again. One reason I feel this way is that I think the Indian Ocean tropical convective forcing is likely to shift east-southeast toward the very warm SSTs north of Australia during the next few weeks.
Of course, as always there are large uncertainties about timing and synoptic details. However, I think it is probable for both the GWO and MJO to be in roughly phase 5 by early January 2008, as I discussed a week ago. The seasonal cycle will also have “its say in this matter”, since climatologically SSTs 29C and greater become covered with tropical convection during the first part of January. So, are we going to have a period of ocean-atmosphere coupling ~120E-180 south of the equator in about 3 weeks (with the onset of the Australian monsoon)? Are we going to see another decent GWO/MJO circuit afterwards? The answers to these types of predictability questions are unclear; however, I speculate yes.
A generally active weather regime is probable to continue for most of the country through at least the end of 2007. As I typed a week ago, the period from around Christmas-early January 2008 may be exceptionally stormy focusing on the Rockies and Plains in the presence of an Arctic cold air source. However, both coasts and Alaska are probable to have impacts from our active regime. I also think it is probable to see strong ridge amplification (greater than climatology) from just west of North America-Alaska by mid January per above reasoning. The latter would suggest an anomalous cold regime from east of the Rockies-east coast with the possible exception of the Deep South.
Internationally, the main tropical cyclone hazard should shift from the South Indian Ocean (Bay of Bengal understood) toward the region of Australia and eventually southwest Pacific Ocean weeks 2-3. Surface westerly wind anomalies ~10-15m/s in that region certainly support this notion. By that time the Australian monsoon should be on-going. Indonesia and even the Philippines are also probable to have possibly severe rainfall events by weeks 2-3. I would think the current Scandinavian blocking pattern should break down by week 3, possibly retrograding into the North Atlantic Ocean. In any case, cool/wet conditions seem probable for much of the Mediterranean region through week-2. Arctic air will continue to build up and expand across particularly Siberia as well as Alaska and much of Canada during the next few weeks. Finally, I hope the spirit of Gabrielle will rest for the warm North Atlantic basin until boreal summer 2008.
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 “convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I may not be able to post another discussion for at least a couple weeks; stay tuned.
Ed Berry
Saturday, December 08, 2007
AAM Crashing (like the recent finacial markets)
Yes, there has been a rebound in the USA stock markets recently especially last week. I hope this trend continues.
There is no change to my reasoning from a week ago; hence some brevity is appropriate for this discussion. The spatial distribution of global tropical SSTs remain essentially the same, with minus 1-3C anomalies along the equator from 160E-west coast of South America. Warmest SSTs across the Indo-Pacific sector are located from the north coast of Australia to Indonesia with totals in excess of 30C in places. Increasing trades from the equatorial date line into the west central Pacific Ocean (anomalies ~5-10m/s) may add some re-invigoration to the on-going El-Viejo.
An oceanic Kelvin wave was generated by the westerly wind burst forced by the recent MJO. Per latest five-day averaged TAO buoy data anomalies in excess of 3C at ~150m depth in the region of the equatorial date line are associated with this Kelvin wave. The impacts of this and additional Kelvin waves onto our basin wide moderate cold event (in terms of SSTs) is a monitoring issue. However, I speculate that the early stages of a transition to El-Nino are in progress, especially given the biennial character of ENSO over the last several years.
The dynamical signal with the MJO is quickly returning to the Eastern Hemisphere. In fact, a back of the envelope calculation from near equatorial velocity potential Hovmoller plots gives me a phase speed of ~30m/s as this signal propagated through the Western Hemisphere. Interactions with the extratropics along with decoupling between the atmosphere and tropical convection allow for these kinds of fast phase speeds as the MJO response moves through the Western Hemisphere.
Full disk satellite imagery shows tropical convective forcing rapidly increasing across the equatorial Indian Ocean during the last several days, with enhanced rainfall still present across portions of tropical South America and South Africa. WH2004 phase space plots support this observation, suggesting that the MJO signal is currently in phase 1. Animations of upper tropospheric daily mean vector wind anomalies not only show twin anticyclones developing in the region of the Indian Ocean, but also easterly wind flow returning to the equatorial atmosphere. In the zonal mean equatorial wind anomalies are already negative, and what is left of the westerly wind flow anomalies has propagated poleward well into the subtropics and midlatitudes of both hemispheres (through eddy feedbacks).
The combination of frictional dissipation, eddy dynamical processes and the tropical forcing returning to the Eastern Hemisphere is rapidly putting the brakes on the global atmosphere (and speeding up the earth shortening the length of day by a few milliseconds) as I type. In other words, easterly wind flow anomalies are quickly being added back to the equatorial and subtropical atmospheres creating a momentum sink. Given our La-Nina base state this circulation response was expected once the MJO returned to the Eastern Hemisphere.
Per ESRL/PSD AAM plots updated through 6 December using the R1 data climatology, 3-day averaged global relative AAM tendency is at least minus 40 Hadleys. The surface torques as well as the Coriolis torque are contributing to this big AAM tendency. Actual global relative AAM will crash very soon, if it is not already doing so. The expected large circuit of the GWO is occurring, currently entering phase 1. During big subseasonal events such as what has been occurring over the last several weeks, both the MJO and GWO are probable to be in similar phases.
My feeling is that another large GWO orbit and respectable MJO are probable during the next ~50 days. Of course, uncertainties about timing, synoptic details, etc., are always an issue. The GWO should circuit to phase 3 (old GSDM Stage 1) by the end of week-2 and perhaps phase 5 (Stage 2) by the end of week-4 (~ 5 January 2008). Thus my USA outlook from a week ago remains the same. I think there will be a bit of a respite from the currently active western/central USA winter regime week-2 as troughs initially deepen along/off the west coast during that period. Most week-2 ensemble means now show this response. While the USA west coast (and Alaska at times) receives several rounds of possibly intense/severe precipitation during week-2, the rest of the country east of the Divide should be milder. More “Kona Lows” may also be probable for Hawaii.
Starting around Christmas and continuing well into January a cold and active regime appears probable especially if the tropical forcing initially couples to the warm SSTs north of Australia. Troughs would again be favored across the western states with an active southwest flow storm track across the Plains. Sparing details, this pattern may shift ~10-20 degrees of longitude farther east during mid-late January bringing possibly intense cold for particularly the northeast states while extending back into the central USA. At that time ridge amplification may occur from just off the USA west coast into Alaska.
Internationally, an increasing tropical cyclone hazard across the South Indian Ocean especially by week-2 is probable. Locations from the Philippines into the west central Pacific should enjoy some quieter weather at least week-1. Cold Arctic air will continue to build up across Siberia as well as Canada during the next several weeks.
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 “convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will try to post another discussion around next weekend or early week-2.
Ed Berry
There is no change to my reasoning from a week ago; hence some brevity is appropriate for this discussion. The spatial distribution of global tropical SSTs remain essentially the same, with minus 1-3C anomalies along the equator from 160E-west coast of South America. Warmest SSTs across the Indo-Pacific sector are located from the north coast of Australia to Indonesia with totals in excess of 30C in places. Increasing trades from the equatorial date line into the west central Pacific Ocean (anomalies ~5-10m/s) may add some re-invigoration to the on-going El-Viejo.
An oceanic Kelvin wave was generated by the westerly wind burst forced by the recent MJO. Per latest five-day averaged TAO buoy data anomalies in excess of 3C at ~150m depth in the region of the equatorial date line are associated with this Kelvin wave. The impacts of this and additional Kelvin waves onto our basin wide moderate cold event (in terms of SSTs) is a monitoring issue. However, I speculate that the early stages of a transition to El-Nino are in progress, especially given the biennial character of ENSO over the last several years.
The dynamical signal with the MJO is quickly returning to the Eastern Hemisphere. In fact, a back of the envelope calculation from near equatorial velocity potential Hovmoller plots gives me a phase speed of ~30m/s as this signal propagated through the Western Hemisphere. Interactions with the extratropics along with decoupling between the atmosphere and tropical convection allow for these kinds of fast phase speeds as the MJO response moves through the Western Hemisphere.
Full disk satellite imagery shows tropical convective forcing rapidly increasing across the equatorial Indian Ocean during the last several days, with enhanced rainfall still present across portions of tropical South America and South Africa. WH2004 phase space plots support this observation, suggesting that the MJO signal is currently in phase 1. Animations of upper tropospheric daily mean vector wind anomalies not only show twin anticyclones developing in the region of the Indian Ocean, but also easterly wind flow returning to the equatorial atmosphere. In the zonal mean equatorial wind anomalies are already negative, and what is left of the westerly wind flow anomalies has propagated poleward well into the subtropics and midlatitudes of both hemispheres (through eddy feedbacks).
The combination of frictional dissipation, eddy dynamical processes and the tropical forcing returning to the Eastern Hemisphere is rapidly putting the brakes on the global atmosphere (and speeding up the earth shortening the length of day by a few milliseconds) as I type. In other words, easterly wind flow anomalies are quickly being added back to the equatorial and subtropical atmospheres creating a momentum sink. Given our La-Nina base state this circulation response was expected once the MJO returned to the Eastern Hemisphere.
Per ESRL/PSD AAM plots updated through 6 December using the R1 data climatology, 3-day averaged global relative AAM tendency is at least minus 40 Hadleys. The surface torques as well as the Coriolis torque are contributing to this big AAM tendency. Actual global relative AAM will crash very soon, if it is not already doing so. The expected large circuit of the GWO is occurring, currently entering phase 1. During big subseasonal events such as what has been occurring over the last several weeks, both the MJO and GWO are probable to be in similar phases.
My feeling is that another large GWO orbit and respectable MJO are probable during the next ~50 days. Of course, uncertainties about timing, synoptic details, etc., are always an issue. The GWO should circuit to phase 3 (old GSDM Stage 1) by the end of week-2 and perhaps phase 5 (Stage 2) by the end of week-4 (~ 5 January 2008). Thus my USA outlook from a week ago remains the same. I think there will be a bit of a respite from the currently active western/central USA winter regime week-2 as troughs initially deepen along/off the west coast during that period. Most week-2 ensemble means now show this response. While the USA west coast (and Alaska at times) receives several rounds of possibly intense/severe precipitation during week-2, the rest of the country east of the Divide should be milder. More “Kona Lows” may also be probable for Hawaii.
Starting around Christmas and continuing well into January a cold and active regime appears probable especially if the tropical forcing initially couples to the warm SSTs north of Australia. Troughs would again be favored across the western states with an active southwest flow storm track across the Plains. Sparing details, this pattern may shift ~10-20 degrees of longitude farther east during mid-late January bringing possibly intense cold for particularly the northeast states while extending back into the central USA. At that time ridge amplification may occur from just off the USA west coast into Alaska.
Internationally, an increasing tropical cyclone hazard across the South Indian Ocean especially by week-2 is probable. Locations from the Philippines into the west central Pacific should enjoy some quieter weather at least week-1. Cold Arctic air will continue to build up across Siberia as well as Canada during the next several weeks.
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 “convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will try to post another discussion around next weekend or early week-2.
Ed Berry
Saturday, December 01, 2007
Subprime Woes for the AAM Budget?
Other titles for this posting might have been, “The Horrors of Making Weather Predictions”, and “Spinning Wheel”. The latter is a song by Blood, Sweat and Tears a number of years ago, and it had a lyric something to the effect of, “what goes up must come down”. The point is we are in a global circulation state that is extremely difficult to understand (and write about), let alone forecast it. In other words, stochastic forcing rules. Hey, yet another title to use for later discussions. I offer below what some would simply view as an opinion (worth something, anyway).
The spatial distribution of equatorial and tropical SSTs across the Indo-Pacific sector remains consistent with a mature La-Nina (a.k.a. a cold event and El-Viejo). Negative anomalies are present in a wide equatorial band (~ +- 10 deg latitude of zero latitude) from 160E-west coast of South America, with anomalies lower than minus 2C in places extending to depths ~150m. Per 5-day averaged TAO buoy data ending 30 November, anomalies even less than minus -3C were observed ~150m/140W. Also noteworthy is that west of 160W a warm anomaly of ~plus 3C has recently developed at 150m. More said below about the possible importance of this feature.
The large scale character of these below normal El-Viejo SSTs was exasperated by a recent trade wind surge. Positive SST anomalies have become well defined from ~20S/180-north of Australia into the equatorial Indian Ocean, the latter a region of convective suppression during the last few weeks. Contributing to these warm SST anomalies has been the southward shift with the seasonal cycle, having SST totals in excess of 30C in places. Much of the equatorial and northern Atlantic Ocean basin remains above average, as does the North Pacific Ocean basin. In fact, there is still some signal of the warm extratropical horseshoe pattern of SSTs across the Pacific Ocean. Recent East Asian jet extensions and associated cold outbreaks and accompanying baroclinic storm developments across the North Pacific have led to cooling centered around the extratropical date line. The latter is an example of the atmosphere forcing the ocean, typical for the extratropics.
A strong signal of the MJO has emerged during the last several days. This was not expected by me, even though various statistical predictions of the WH2004 phase space diagrams did indicate this possibility. One reason I disagreed (and there were others) is that it is somewhat atypical during a moderate-strong La-Nina to have a coherent eastward propagating MJO globally. I should not be surprised given all the other “craziness” I have had to discuss in these postings during the last couple years. For example, behaviors I have termed as “The New World Atmosphere” and “Nemesis”.
The eastward propagation initiated ~80E along the equator during about mid-October. An intense tropical convective flare-up occurred ~140-160E during roughly 9-26 November, associated with the MJO signal along with extratropical forcing. Currently, the moist convective signal is centered ~15S/160W, but with other convection increasing across tropical South America and Africa. Full disk satellite imagery and coherent modes Hovmollers support this observation. My back of the envelop phase speed computation gives me ~3-4m/s eastward movement for the convective signal during about the last 45 days. Interestingly, there was a weaker but significant tropical convective flare-up ~150E during mid-late September, which is also on the time scale of ~50-60 days. The point is perhaps greater attention should have been paid to those “New World Atmosphere” SSTs across the west central Pacific Ocean.
The strong MJO projection onto the WH2004 RMM phase space plots has been coming from the second EOF, RMM2. Contributions to this ~3 standard deviation signal have been from the convectively suppressed Indian Ocean (positive OLRA) along with the equatorial vertical wind structure. The statistical RMM tools indicate the MJO dynamical signal to return into the equatorial Indian Ocean by the end of week 3. Hey, I have to go with it.
Now comes the portion discussing the AAM budget and GWO. The earth-atmosphere budget is about as complicated as it gets; however, I do think I can make some sense of it. I want to be relative brief about it.
Tied to the ~160E tropical flare-up discussed above along with a largely Southern Hemisphere forced friction-mountain torque (yes, the Northern Hemisphere also helped out) index cycle, a non-trivial amount of westerly wind flow was added to the atmosphere. Since early November ~3 AMUs was added to the global circulation, focusing not only across the equatorial regions, but also the northern and southern midlatitudes, in terms of the zonal mean. The midlatitude contribution was a response to eddy feedback dynamics that much research still needs to be done to understand it. There has been a recent peak of global relative AAM to slightly above the R1 data climatology. Just as the low AAM base state observed much of boreal autumn was an extreme weather event, this sudden increase was also the same.
The GWO presents this signal nicely, very weakly in the phase 7 (legacy GSDM Stage 3) plane through 28 November (5-day averaged). Even though the GWO projection is near zero, the variation is also a 2-3 standard deviation event. So, we have experienced 2-3 sigma events of both the GWO and MJO. The GWO signal can be seen in terms of broad zonal mean westerly wind anomalies from ~30N-30S with magnitudes ~5m/s. A true MJO would be more equatorially confined before propagating poleward. The punch line to all this will be to see if global relative AAM goes down as fast as it came up during the next few weeks. Now you know where analogies to the recent subprime woes of the financial markets come in, etc.
Currently the global surface torques have become weakly negative (through 29 November) as seen by lowering mean sea level pressures along north-south mountain ranges and anomalous surface westerly wind flow across the tropical and extratropical Pacific Ocean. The latter represents frictional dissipation of the westerly flow. In fact, a strong westerly wind burst at ~150E accompanied the MJO convection discussed above, and I think a down welling oceanic Kelvin wave may have been initiated leading to the subsurface warming also mentioned above.
The AAM transport signal is incredibly complicated. However, there is a decent poleward transport signal at centered ~45N and 50S (roughly 10 Hadleys for Northern Hemisphere). There is evidence that these transports (flux divergence of AAM transport) are leading to meriodional propagation of zonal mean zonal wind anomalies. For instance, zonal mean anomalous easterlies (tied to La-Nina) are shifting into the midlatitude atmospheres while the equatorial westerlies also come off. The midlatitude westerly anomalies are also shifting poleward. Again, there is strong inter-hemispheric symmetry of both zonal mean and regional-scale circulation anomalies due to the strong tropical forcing.
Cutting to the chase of this mess, should the tropical convective forcing shift into the Indian Ocean/Maritime Continent during weeks 2-3, while anomalous ridging may be favored at the higher latitudes, anomalous westerly wind flow may be probable from the subtropics-lower midlatitudes. The latter would be a response to anomalous easterly wind flow anomalies returning to the equatorial regions of the atmosphere (more said below).
Animations of fields such as upper tropospheric daily mean vector wind anomalies present one of the best Rossby wave energy dispersions (RWD) across the PNA sector I have seen in a while. Linked to twin subtropical anticyclones along/west of the date line, a low-high-low-high pressure great circle arc including the deep western USA trough can be seen 1 December 2007. A large anticyclone is also bursting into the Arctic, which could help build up more bitterly cold air. Wind anomalies in excess of 50m/s have occurred with this RWD during the last several days. This situation serves as an example of an extreme synoptic weather event linked to the slower processes involving both the GWO and MJO per above. Another point is that we have experienced an extreme event of subseasonal atmospheric variability that already has and will impact multiple time scales until further notice. For instance, are we beginning the slow process of transitioning from El-Viejo to El-Nino?
Let’s move onto insights for weeks 1-4. I can only wish for the day others can observe and appreciate the variations from the GWO (GSDM) perspective that we try to share. As stated above, phase 7 of the GWO best describes the current weather-climate situation. Any predictive information I offer here is likely to going to be “wrong” anyway (low confidence, seriously), so I may as well “go for it”. Again, stochastic forcing rules, and any objective prediction schemes will likely suffer more than usual during the next few weeks.
Most models already suggest some form of amplification weeks 1-2, including the possibility of Plains storm development ~10-14 December. This is a response to the likely collapse of the strong North Pacific jet. In any case, my feeling would be expect a situation of a deep east Asian trough-central/east Pacific ridge (into Alaska and polar latitudes at times) leading to at times an anomalous trough focused across the western/central USA by weeks 3-4. That option would be an opinion to respect should AAM crash and the GWO strongly circuits to ~phase 3 (old GSDM Stage 1) particularly if westerly flow remains strong across the subtropics. Ramifications for the USA may be sporadic episodes of “winter” weeks 1-2 (Pacific Northwest wetness, cold across the northern states, etc), followed by a significant cold and wet regime weeks 3-4. The most intense cold and storm track activity by that time may focus on the Rockies and Plains; however, locations such as the northeast states may be impacted as well. I think other USA concerns are apparent.
Internationally, in addition to an increasing tropical cyclone hazard across the South Indian Ocean weeks 1-2, there should be some concern for this hazard across the South Pacific Ocean paradise islands as well. Much of Europe is likely to remain active while portions of tropical South America get beneficial rainfall. Finally, at least seasonable cold is probable for Siberia (to come into the USA by around Christmas?).
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 "convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will try to post another discussion around next weekend or early week-2.
Ed Berry
The spatial distribution of equatorial and tropical SSTs across the Indo-Pacific sector remains consistent with a mature La-Nina (a.k.a. a cold event and El-Viejo). Negative anomalies are present in a wide equatorial band (~ +- 10 deg latitude of zero latitude) from 160E-west coast of South America, with anomalies lower than minus 2C in places extending to depths ~150m. Per 5-day averaged TAO buoy data ending 30 November, anomalies even less than minus -3C were observed ~150m/140W. Also noteworthy is that west of 160W a warm anomaly of ~plus 3C has recently developed at 150m. More said below about the possible importance of this feature.
The large scale character of these below normal El-Viejo SSTs was exasperated by a recent trade wind surge. Positive SST anomalies have become well defined from ~20S/180-north of Australia into the equatorial Indian Ocean, the latter a region of convective suppression during the last few weeks. Contributing to these warm SST anomalies has been the southward shift with the seasonal cycle, having SST totals in excess of 30C in places. Much of the equatorial and northern Atlantic Ocean basin remains above average, as does the North Pacific Ocean basin. In fact, there is still some signal of the warm extratropical horseshoe pattern of SSTs across the Pacific Ocean. Recent East Asian jet extensions and associated cold outbreaks and accompanying baroclinic storm developments across the North Pacific have led to cooling centered around the extratropical date line. The latter is an example of the atmosphere forcing the ocean, typical for the extratropics.
A strong signal of the MJO has emerged during the last several days. This was not expected by me, even though various statistical predictions of the WH2004 phase space diagrams did indicate this possibility. One reason I disagreed (and there were others) is that it is somewhat atypical during a moderate-strong La-Nina to have a coherent eastward propagating MJO globally. I should not be surprised given all the other “craziness” I have had to discuss in these postings during the last couple years. For example, behaviors I have termed as “The New World Atmosphere” and “Nemesis”.
The eastward propagation initiated ~80E along the equator during about mid-October. An intense tropical convective flare-up occurred ~140-160E during roughly 9-26 November, associated with the MJO signal along with extratropical forcing. Currently, the moist convective signal is centered ~15S/160W, but with other convection increasing across tropical South America and Africa. Full disk satellite imagery and coherent modes Hovmollers support this observation. My back of the envelop phase speed computation gives me ~3-4m/s eastward movement for the convective signal during about the last 45 days. Interestingly, there was a weaker but significant tropical convective flare-up ~150E during mid-late September, which is also on the time scale of ~50-60 days. The point is perhaps greater attention should have been paid to those “New World Atmosphere” SSTs across the west central Pacific Ocean.
The strong MJO projection onto the WH2004 RMM phase space plots has been coming from the second EOF, RMM2. Contributions to this ~3 standard deviation signal have been from the convectively suppressed Indian Ocean (positive OLRA) along with the equatorial vertical wind structure. The statistical RMM tools indicate the MJO dynamical signal to return into the equatorial Indian Ocean by the end of week 3. Hey, I have to go with it.
Now comes the portion discussing the AAM budget and GWO. The earth-atmosphere budget is about as complicated as it gets; however, I do think I can make some sense of it. I want to be relative brief about it.
Tied to the ~160E tropical flare-up discussed above along with a largely Southern Hemisphere forced friction-mountain torque (yes, the Northern Hemisphere also helped out) index cycle, a non-trivial amount of westerly wind flow was added to the atmosphere. Since early November ~3 AMUs was added to the global circulation, focusing not only across the equatorial regions, but also the northern and southern midlatitudes, in terms of the zonal mean. The midlatitude contribution was a response to eddy feedback dynamics that much research still needs to be done to understand it. There has been a recent peak of global relative AAM to slightly above the R1 data climatology. Just as the low AAM base state observed much of boreal autumn was an extreme weather event, this sudden increase was also the same.
The GWO presents this signal nicely, very weakly in the phase 7 (legacy GSDM Stage 3) plane through 28 November (5-day averaged). Even though the GWO projection is near zero, the variation is also a 2-3 standard deviation event. So, we have experienced 2-3 sigma events of both the GWO and MJO. The GWO signal can be seen in terms of broad zonal mean westerly wind anomalies from ~30N-30S with magnitudes ~5m/s. A true MJO would be more equatorially confined before propagating poleward. The punch line to all this will be to see if global relative AAM goes down as fast as it came up during the next few weeks. Now you know where analogies to the recent subprime woes of the financial markets come in, etc.
Currently the global surface torques have become weakly negative (through 29 November) as seen by lowering mean sea level pressures along north-south mountain ranges and anomalous surface westerly wind flow across the tropical and extratropical Pacific Ocean. The latter represents frictional dissipation of the westerly flow. In fact, a strong westerly wind burst at ~150E accompanied the MJO convection discussed above, and I think a down welling oceanic Kelvin wave may have been initiated leading to the subsurface warming also mentioned above.
The AAM transport signal is incredibly complicated. However, there is a decent poleward transport signal at centered ~45N and 50S (roughly 10 Hadleys for Northern Hemisphere). There is evidence that these transports (flux divergence of AAM transport) are leading to meriodional propagation of zonal mean zonal wind anomalies. For instance, zonal mean anomalous easterlies (tied to La-Nina) are shifting into the midlatitude atmospheres while the equatorial westerlies also come off. The midlatitude westerly anomalies are also shifting poleward. Again, there is strong inter-hemispheric symmetry of both zonal mean and regional-scale circulation anomalies due to the strong tropical forcing.
Cutting to the chase of this mess, should the tropical convective forcing shift into the Indian Ocean/Maritime Continent during weeks 2-3, while anomalous ridging may be favored at the higher latitudes, anomalous westerly wind flow may be probable from the subtropics-lower midlatitudes. The latter would be a response to anomalous easterly wind flow anomalies returning to the equatorial regions of the atmosphere (more said below).
Animations of fields such as upper tropospheric daily mean vector wind anomalies present one of the best Rossby wave energy dispersions (RWD) across the PNA sector I have seen in a while. Linked to twin subtropical anticyclones along/west of the date line, a low-high-low-high pressure great circle arc including the deep western USA trough can be seen 1 December 2007. A large anticyclone is also bursting into the Arctic, which could help build up more bitterly cold air. Wind anomalies in excess of 50m/s have occurred with this RWD during the last several days. This situation serves as an example of an extreme synoptic weather event linked to the slower processes involving both the GWO and MJO per above. Another point is that we have experienced an extreme event of subseasonal atmospheric variability that already has and will impact multiple time scales until further notice. For instance, are we beginning the slow process of transitioning from El-Viejo to El-Nino?
Let’s move onto insights for weeks 1-4. I can only wish for the day others can observe and appreciate the variations from the GWO (GSDM) perspective that we try to share. As stated above, phase 7 of the GWO best describes the current weather-climate situation. Any predictive information I offer here is likely to going to be “wrong” anyway (low confidence, seriously), so I may as well “go for it”. Again, stochastic forcing rules, and any objective prediction schemes will likely suffer more than usual during the next few weeks.
Most models already suggest some form of amplification weeks 1-2, including the possibility of Plains storm development ~10-14 December. This is a response to the likely collapse of the strong North Pacific jet. In any case, my feeling would be expect a situation of a deep east Asian trough-central/east Pacific ridge (into Alaska and polar latitudes at times) leading to at times an anomalous trough focused across the western/central USA by weeks 3-4. That option would be an opinion to respect should AAM crash and the GWO strongly circuits to ~phase 3 (old GSDM Stage 1) particularly if westerly flow remains strong across the subtropics. Ramifications for the USA may be sporadic episodes of “winter” weeks 1-2 (Pacific Northwest wetness, cold across the northern states, etc), followed by a significant cold and wet regime weeks 3-4. The most intense cold and storm track activity by that time may focus on the Rockies and Plains; however, locations such as the northeast states may be impacted as well. I think other USA concerns are apparent.
Internationally, in addition to an increasing tropical cyclone hazard across the South Indian Ocean weeks 1-2, there should be some concern for this hazard across the South Pacific Ocean paradise islands as well. Much of Europe is likely to remain active while portions of tropical South America get beneficial rainfall. Finally, at least seasonable cold is probable for Siberia (to come into the USA by around Christmas?).
Appendix
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
http://www.cdc.noaa.gov/map/images/gcm/gsdm_95d.jpg
http://www.cdc.noaa.gov/map/images/gcm/gsdm_40d.jpg
http://www.cdc.noaa.gov/map/clim/gsdm.shtml
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the legacy GSDM is to extend current thinking beyond the MJO, the GWO quantifies variations used to derive the original GSDM in a manner that is “user friendly” analogous the WH2004 "convention”.
Links to CPC and PSD ENSO discussions:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/index.shtml
http://www.cdc.noaa.gov/people/klaus.wolter/MEI/
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us (soon) to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will try to post another discussion around next weekend or early week-2.
Ed Berry
Subscribe to:
Posts (Atom)