« Previous | Main

Formula One Weather Forecast: Brazilian Grand Prix 2010

Ian Fergusson | 08:01 UK time, Tuesday, 2 November 2010

Interlagos, 5-7 November 2010 (Round 18)

(This forecast will be regularly updated. This entry: Friday 5 November, 15:00hrs GMT)

You can follow me on Twitter to get regular forecast updates, too

Quick Links:


 Forecast Summary:

(PPN = Precipitation)

Emphasis: Increasing confidence sees Friday to remain dry or largely so - possible arrival of some afternoon showers aside - before a thundery trough/cold front moving north brings threat of wet weather overnight into Saturday AM. The remainder of Saturday sees further thunderstorms (possibly severe, with torrential downpours in places), showers or prolonged spells of rain likely during the morning; still a threat of showers (some heavy) continuing into the FP3/quali windows. Currently good inter-model agreement for Sunday to be dry or largely so for the race, with a fairly low % chance of some showers forming later PM.

Friday: CONFIDENCE HIGH: Sunny spells and dry for much of the day, before cloud increases later PM, bringing showers/t-storms possibly by late afternoon but more especially into the evening and then overnight. Chance of PPN 25% until end of FP2. Hot; Max 29C.  Breezier later. Wind NNW-NNE

Saturday: CONFIDENCE MODERATE-HIGH (PPN totals subject to some doubt): Cloudy. Outbreaks of rain or showers/t-storms arriving overnight and continuing into the morning, heavy at times, some severe t-storms possible.  A continued periodic threat of some further heavy showers / spells of rain appearing throughout the afternoon, yielding some very wet conditions on track. Chance of PPN 90%. Cooler. Max 21C. Wind light-moderate, variable; mostly S.

Sunday: CONFIDENCE HIGH: Perhaps more widespread early cloud, but this breaking to offer sunny spells for remainder of the day. Some deeper convective cloud build-up is likely later into the afternoon with a few (mostly light) showers possible, but the balance currently firmly favours dry conditions. Chance of PPN 25%.  Max 23C.  Wind light-moderate, SE-ESE. 


Synopsis & Forecast Evolution:  

 Fri. Update, based on 00 & 06UTC model output:

Pretty good inter-model agreement now for the next 24-36hrs, with the cold front and associated thundery band now very evident on IR satellite loop (see link above) steadily approaching from the SW. The resultant rainfall later Friday overnight and into Saturday has potential to be heavy at times, albeit the actual extremes of PPN accumulation remain open to uncertainty. Recent runs of GFS have turned to become one of the more extreme solutions, falling somewhat closer into line with earlier output throughout the medium range from most other operational centres. However, the variation in rainfall modelling between models still needs emphasising; e.g., through the next 24hrs, Brazilian global output (RSPAS) offers fairly conservative rain totals across much of São Paulo (<5mm), while their regional 20-km resolution model (ETA-20) concurs with GFS and some others (including UKMO and ECMWF) on significant accumulations of between 40-100mm. As explained in my update yesterday,  there's ample potential for mass convergence and organisation of the thundery zone during it's northward passage and the potential for severe weather certainly cannot be discounted during Friday night and Saturday. Could we see a repeat of scenes from last year's event? Very possible, but timing will be critical.

Meanwhile, the models still hold onto the notion of a dry (or drier) affair on Sunday. The upper trough is expected to lose intensity, albeit some models vary their handling of prevailing flow at mid and upper levels, thus reaching different conclusions of the position and amplitude/phasing of the trough and associated PPN. Current consensus sees the rain further to the NE, with moisture convergence activated across the Amazon and across to the Atlantic coasts, yielding thunderstorms once again but not affecting São Paulo. Still subject to some flux, for sure, but for now the prospect of a dry race remaiins the stronger bet.


Thurs. Update, based on 00 & 12UTC model output:

There's growing inter-model and inter-run continuity now, with little significant to add to the general theme described since the outset of this blog. Much of Friday will remain dry and fair, but with deeper convective development into the afternoon (perhaps even by FP2) and evening, heralding the arrival of the cold front from the south. This becomes a dominant feature overnight into Saturday, with most models tending towards clearance (of the main PPN band) by around midday on Saturday, but the finer detail of local rainfall threat shouldn't be taken too literally, given the varied handling of frontal PPN spatio-temporal extent by different models. Some models - notably ECMWF, UKMO, CMC and Brazilian regional NWP output - have been bullish about the wet weather potential for Saturday throughout the medium range lead-up. Much like we saw in Korea, GFS has somewhat turned around to fall more into line with these solutions, but with inevitable run-to-run oscillations re finer detail. I expect this to remain so right up to the wire: after all, ignoring the mesoscale continuity, a dozen miles will make a huge difference in terms of how the thundery band clears (or not) ahead of the track sessions on Saturday, thus dictating fortunes on the circuit. Radars at the ready, folks...

There's a cocktail of critical upper and low-level atmospheric dynamics coming into play, including potential for localised mass convergence; import of moist air via low-level jet (LLJ) out of the vast Amazon basin and consequently, a prospect of seeing a more organised structure of any frontal/trough passage, offering a risk of heavy downpours in places. The fairly low pre-frontal pressure gradient is significant, however. GFS (and some other models) still prefer a somewhat discontinuous feature from Atlantic into the continental interior, making the local PPN accumulation even harder to reliably call, whereas others (including UKMO) suggest a more coherent band of heavy rain, with slower rate of clearance during Saturday. Of relevance here is how regional Brazilian modellling for Sao Paulo (i.e., ETA20, ETA40) tend to concur with UKMO and ECMWF in threatening higher amounts of PPN, in contrast to some other global output (e.g., GFS) - this all courtesy of differences in how the upper level flow is synthesised.  

 Either way, the potential for heavy rain (plus low but no less emphasised threat of severe t-storms) at some stage early Saturday comes with high forecast confidence: a wet, damp or drying track being the likely outcome for FP3 and possibly quali too. Any cloud breaks PM will add insolation effects and bolster a threat of further localised showers, hence the % chance of some rain never really diminishes to any significant degree during the periods of track action. It could prove knife-edge, not least given the latent drama surrounding this race and it's importance for the championship tables...

Sunday still looks a far better affair, with a goood deal of dry and largely sunny weather prevailing and a broadly low  but not insignificant threat of afternoon light showers (as exemplified by both UKMO and ECMWF output)). Further model runs are needed to lock-down the likely chance of these, but at least broad inter-model consensus offers a high chance for a dry or largely dry race.

Weds. Update, based on 00UTC model output:

A largely similar story to my original entry and some better continuity - of sorts - is now emerging. The broadscale evolution has good inter-model agreement, with the upper trough dominant from overnight Friday and then throughout Saturday; a cold front moving north later Friday overnight into Saturday AM. Inevitably, each model handles the precipitation signal somewhat differently, i.e., the areal extent of showers and timing of frontal clearance northwards. In short, no model can be wholly trusted on this level of detail, not least at this range in a convective regime. Importantly, the new runs of ECMWF, UKMO and CMC are a tad more progressive (albeit not by much) in easing the frontal band of showers and t-storms northwards through Saturday morning to possibly clear Interlagos by midday (or indeed earlier). GFS still remains the most progressive in this respect, with good continuity; similarly Eta-20 regional modelling now follows it's lead. It also has the primary PPN signal restricted to overnight Fri-Sat, followed by further late afternoon/evening showers on a trough post-front. This sort of shortwave feature (plus any local diurnal shower development) inevitably keeps a threat of some showers still present into the afternoon, but there's a fair chance of the wettest conditions having passed ahead of FP3 and quali, perhaps by some margin. Sunday, meanwhile, continues to look dry or very largely so, with good inter-model support for this outcome (as earlier noted) and a dry race expected.


(original entry follows) 

So, are we ready for the drama? Interlagos seems to have an annual ability to offer heart-stopping moments - not least in the last three championship-deciding events there.  And if there's a last weather-borne banana skin to await the 2010 title contenders, it will surely appear here at São Paulo, rather than at Abu Dhabi...

The potential is certainly there for Brazil's lively spring weather to play some part in the destiny of this year's title hopes. The key focus is on Saturday, when the possibility of (at least periodically) wet weather will be of most concern for teams and drivers as the early forecast evolves.

The key models employed for this forecast blog are the global models from UK MetOffice (UKMO); ECMWF, US-GFS, Canada's CMC, Japan's JMA, US-NOGAPS and the regional (20km resolution) output from Brazil's INPE/CPTEC.

During the later stages of Friday, an upper trough elongates northwards out of Argentina, with a surface low spinning-up offshore to the east. Up aloft, the polar jetstream winds loop northwards to define the upper trough axis, offering potential for inclement weather to gather below. The subtropical jet lies just to the north. Showers and thunderstorms are likely to become increasingly more prevalent beneath this vigorous cocktail of upper forcing, aided locally by daytime maxima climbing into the mid/upper 20's.

The US-GFS is not wholly dissimilar, offering dry conditions for much of Friday before also bringing the upper trough across São Paulo later into Saturday, albeit as a somewhat less pronounced and flatter feature compared to some other models.   

As the weekend advances, the trough expands up through southern Brazil into Saturday, before relaxing eastwards into Sunday. Meanwhile, just to the north, the SACZ (South Atlantic Convergence Zone - a critical component of Brazilian weather patterns) sits almost quasi-stationary, delivering frequent thunderstorms stretching over a vast band from the depths of the Amazon basin down towards SE Brazil. The intensity and positioning of the SACZ - shown to be allied to phasing of the Madden-Julian Oscillation (MJO) and El Niño Southern Oscillation (ENSO) events - is the most important driving force behind any extreme spring-summer rainfall events occuring in São Paulo.

Friday looks largely or entirely dry during the two free practice sessions, with sunny spells and turning somewhat breezier into the afternoon. Model continuity for this outcome has been essentially consistent throughout the medium range output from all operational centres. This dry spell precedes the increasing influence of the upper trough and surface cold front easing northwards into Brazil later into the evening and night. Surface pressure steadily falls, the cloud cover thicken-up and heralds a streadily increasing % chance of showers later.

At the time of writing, the most pessemistic forecasts for Saturday come from the very same models that performed so admirably when foretelling of inclement conditions a fortnight ago in Korea.

ECMWF, UKMO and CMC are largely in agreement (albeit with nuances of timing differences) in developing a sharper, thundery feature along the frontal band during early hours and keeping it largely in-situ through the day (almost a component of the SACZ), with resultant potential for more prolonged outbreaks of rain at times and generally a lot more cloud across Interlagos. Temperatures will fall back towards the mid 20's. It's noteworthy how the wetter signal for Saturday is also reflected in the Brazilian Eta-20 regional modelling, especially into the afternoon (see latest São Paulo meteogram, link at the top of this blog), thus offering pretty good support for the EC, UKMO and CMC solutions.  ECMWF's EPSgram for São Paulo (NB: not available publically on the web) is a fairly decent replica of the Eta-20 solution, too.

GFS, on the other hand, is more progressive and takes the frontal band northwards quicker (it also wants the SACZ further north, too), with the shower threat easing away before qualifying and skies tending to clear somewhat. Of course, taking the precipitation distribution from GFS (or any other model) at face value in such a convective set-up is a mug's game, because they're only ever providing a broad brush signal for the regional & local potential, not the actuality!

So, adopting the necessary 'wide-angle' view across all the various model output, very clearly the likelihood of rain at some stage during Saturday is very high, but the comparative speed of this clearing through Saturday is the critical element to dictate wet weather potential into the windows of FP3 and qualifying. Some mainstream forecast providers (e.g., MeteoGroup) offer up to 99% precipitation chance for Saturday, which seems quite reasonable taken across the 24-hr window. However, for periods of track action it may prove closer to 20-40% in terms of site-specific probability. The outcome? Anyone's guess!! Better inter-model continuity is awaited...

But either way, the models do then tend to converge in offering Sunday's dry or largely dry prospects. The upper trough relaxes to a more zonal pattern aloft; the surface front/trough is by then further north and swallowed-up into the broader SACZ convective band over SE Brazil. Later into the afternoon, clouds will tend to bubble-up more readily across Interlagos, aided by insolation and a reasonable amount of CAPE, but any resultant showers are likely to be lighter affairs and more broadly scattered. So, there's more than a fighting chance for this year's race to remain a dry affair.

I'll of course keep you updated on how the model runs this week shape-up prospects for what promises to be a truly thrilling event - wet or dry!





  • Comment number 1.

    Ian, at this admittedly very early stages, what are the chances looking at the initial outlook of a delayed quali/Red Flag conditions like last year?

  • Comment number 2.

    Well, it'd certainly be interesting if Saturday was wet, but Friday and Sunday were dry. Not too sure I'd want a wet quali session to be honest, as at this stage of the season, and the way the season has gone, I'd like to see........how should I put this.........weather that's basically fair to everyone, ie dry weather. If Saturday is wet, it'll make things interesting for the teams if Sunday at that time is still forecast to be dry, as far as setup for quali/race goes. Does look like Saturday (and to a lesser extent Sunday) could be hit and miss though in terms of timing/position of the rain.

  • Comment number 3.

    Hi D_M_N, gosh, it's far too early to say because of almost certain positional/timing errors on the frontal band at this range of forecasting. Some models offer precipitation rates of 2-4mm p/hr, so nothing too dramatic, but I stress the difficulty of this one given the potential for local deep convection / t-storms and associated sudden heavier downpours and much higher PPN rates. Different set-up to last year, however. Note also that the US-GFS is more bullish in taking the primary threat away north before quali, but of course - like any NWP output - it should never be viewed or trusted in isolation. The Brazilian (and some other) models offer a more coherent feature looming over the region, with potential for persistent rain at times (UKMO likewise, hence it driving the 'heavy rain' symbol currently on the BBC Weather web spot forecast). GFS could, of course, be correct in the final outcome, but then (as they were in Korea) so could the distinctly wetter story on Saturday offered collectively at present by ECMWF, UKMO, CMC and ditto the Brazilian output. I'm certainly not fretting about red flags or suchlike at this juncture, because the threat of any rain is by no means a 'done deal', at least in the FP3/Quali windows themselves. But fore sure, the potential is real enough and the teams will be well aware of it from their own forecast bods. Cheers, Ian

  • Comment number 4.

    Ah, OK... fair enough! Thought I'd ask anyways!

  • Comment number 5.

    Hello, Ian.
    I live in Sao Paulo, local race, I love weather and I can speak perfectly the climate of Sao Paulo which is pretty crazy, really at this time of spring / summer few minutes, the time goes from sunny to rainy with trovoadas.A climate similar to that of Malaysia, the late-afternoon thunderstorms, can always be severas.Wait to see what can happen in this race, with this cold front that may come.

    Here are some links that will help you and blog readers:
    [Unsuitable/Broken URL removed by Moderator]

    Congratulations, the blog and coverage of Formula 1 racing.
    Sorry,for my bad english.

  • Comment number 6.

    Hi Ian.

    GFS seems all over the place atm, I'm sure you guys at the MO have been watching the low pressure due over us in the UK next week ;)

    I'd definately say the chance of rain on Sunday is much higher than 10% though. At the very least in the 30-40% range.


  • Comment number 7.

    Hi SpeedRacer, yes the developments next week here could prove bothersome. The GFS variance from other models re Brazil is essentially how the cold front / trough evolution is handled continentally; ie a weaker feature not supported well in other models. Brazilian Met rightly note prospect for low level convergence and influence of low jet, all precursors to more organisation (poss severe storms) as signalled in their own model. The Sunday PPN chance may indeed be upgraded on sight of 00z runs as forcing still looks decent in various models... I concur with you. Cheers, Ian

  • Comment number 8.

    Hello Ian,

    I absolutely love Brazil. It always throws up a magnificent race. However, with the rain that has hit the racetrack in recent years, could this be another abandoned Qualifying?

  • Comment number 9.

    I'm starting to think that Brazil and predicting weather is like seeing a flying pig.

    Predicting weather for that country seems a near on impossibility for this time of the year! I don't think we'll have a proper 'forecast' for Saturday and Sunday until late on Thursday or Friday. Depending on the PPN level, I imagine we could see a nowcasting forecast for both days!

  • Comment number 10.

    @Ian Fergusson BBC Weather

    What time is the weather forecast valid for on the BBC F1 page? For instance the weather graphic on the BBC F1 page at the bottom shows rain for Friday and Saturday but is the forecast provided what is expected at 2pm or 12pm or just the most likely weather for the whole of the day?

    In the future could the Friday,Saturday and Sunday forecast displayed on the BBC F1 page be adjusted to show what weather is expected at the time halfway during the qualifying and halfway during the race so people have a idea of what to expect during the race?

    If there is heavy rain does it also increase the chance of thunderstorms?

    Thanks. Great to see a good weather blog for the F1.

  • Comment number 11.

    Should be an interesting weekend weatherwise, especially Saturday........seems like the weather will throw it's final curveball at the teams for the season. Going to be very interesting to see how all 3 days develop though.

  • Comment number 12.

    so weather-wise, it'll be a 2009 qualy with a 2008 race? mmm... bring it on!

  • Comment number 13.

    Hi F1
    Just re your query regarding the BBC Weather website's graphics and how they represent the forecast conditions:
    Firstly, worth noting that unlike the vast majority of mainstream web forecasts, ours are driven by the UK Met Office's Global Model, not the US-GFS used for most. Consequently, there will be times when it could show a divergent solution to other sites (depending on model agreement or not!). If you scroll down to the weather map lower down the forecast webpage, you can also see a pretty good representation of how the UKMO model offers various parameters including rainfall, surface pressure and temperature. This data feed is driven directly from each global model run at UKMO and makes for handy comparison to the likes of GFS, etc. It's then tweaked for each global region by a duty forecaster at BBC Weather Centre.
    The daily/nightly spot graphics, meanwhile, display the most significant weather type for the day -i.e., throughout the daylight hours - but as we get into near-range timing, you can then expand that day's graphics (by clicking on the arrow at right for the day). This then offers a representation of how the UKMO-GM develops the forecast trend through each phase of the day.
    Like all web-based forecasts using a global model, be it UKMO, GFS or whatever, the location forecasts are driven by what's called GRIB data feed, i.e., a geographic block division using a forecast grid, into which each location falls. This tends to broad-brush the localised detail, obviously, but it's generally a reasonable steer. Of course, trying to offer a shower chance for location X in time window Y is never going to come with 100% accuracy!!
    It's been noteworthy how the UKMO modelling, as represented through the BBC graphics, has remained steadfast all week on the high probability of wet weather during Saturday. It also performed very well at the Korean event and some of us from this end (BBC Weather) can (and do) manually tweak the graphics yet further as needed, based on a consensual approach from the Met Office folks at Exeter. Of course, relying on just one model (such as GFS) for producing such forecasts is inevitably prone to error (i.e., single model output is taken too literally and myopically). UKMO's staff compare and contrast their own model output routinely with that produced by ECMWF, GFS, JMA, Arpege, CMC and others, tweaking then as seems appropriate to reach the best consensual solution... in other words, applying good, flexible science, rather than just being partisan!
    Hope this helps -
    Best, Ian

  • Comment number 14.

    @Ian Fergusson BBC Weather

    Thanks for such a great reply. You should keep it as a link for any future questions on the same subject. It would be good if whoever runs the BBC F1 site rewrites the link to say click here for a 24 hour breakdown of the weather and maybe this reply can be in a FAQ section if people are interested in it.

    You mention the GRIB data, is it uniform blocks equally spaced around the World or do you take into account coastal areas and other land or sea features that can affect local weather conditions?

  • Comment number 15.

    Well, it's currently raining there apparently, as well as some thunder and lightning. I'll be honest, I fear we could see either a cancelled session like Japan, or severely delayed like Brazil 09.

  • Comment number 16.

    I've just heard it is raining. Maybe with rain Button or Hamilton might have a chance if a good quali. However it is to quick to speak as we might have a delay like Korea, or Sunday quali like Japan.

  • Comment number 17.

    Thanks for all your efforts on the F1 forecasts. Hope to hear from you again in 2011, and in the meantime, I'll keep listening on BBC Gloucestershire and watching on Points West.

  • Comment number 18.

    Great to see a good weather blog for the F1.

  • Comment number 19.

    It seems the Brazilian weather are quite predictable, how come the Met office always get the UK weather forecast wrong?


BBC © 2014 The BBC is not responsible for the content of external sites. Read more.

This page is best viewed in an up-to-date web browser with style sheets (CSS) enabled. While you will be able to view the content of this page in your current browser, you will not be able to get the full visual experience. Please consider upgrading your browser software or enabling style sheets (CSS) if you are able to do so.