The Curious Case of Santa Ana

Wherever you fly, you’ll likely find there’s local weather knowledge that’s important to pilots that may be unknown to pilots from other areas. In the Los Angeles basin, Santa Ana winds and the marine layer (aka June Gloom) are two biggies.

Santa Ana winds are dry strong offshore winds that effect Southern California mostly in the  Fall and Winter, but can happen any time of the year. As a perfect example, we’re here in the beginning of May 2013 and getting a classic (and dangerous) Santa Ana wind event.

Usually Santa Ana winds are forecast well in advance, but exactly where they hit and their intensity is subject to significant local variation and can lead to some very strange situations. Let’s examine some METARs from this morning for a few local airports.

Let’s start with Ontario, since they have an observer there 24 hours a day that will make the reports more reliable:

KONT 021153Z 34006KT 6SM BR CLR 13/11 A3008

Here we have a classic report for pre-dawn hours in an inland area during marine layer influence – high humidity (only a 2C dewpoint/temperature spread), light winds (though a bit unusual to be out of the NW), mist and worse than normal visibility. Now, look what happens less than half an hour later as the Santa Ana winds begin.

KONT 021224Z 06019G28KT 9SM CLR 24/M18 A3007 
RMK PK WND 06028/1223 WSHFT 1205

WOW! In that time, we’ve gone from 88% to 5% relative humidity. The temperature has jumped 11C (20F). The winds have shifted direction by 80 degrees and are now gusting up to 28kts with an improvement in visibility. You’d be right to expect a wild ride if you were trying to fly as this happened. Now, let’s see just how local these effects can be and how automated weather reports could leave a pilot really confused.

Chino (CNO) airport is only 5 miles away from Ontario (mostly south and slightly west). As you might expect, it shows a similar trend in winds and humidity, but a few oddities as well for the automated report at about the same time:

KCNO 021237Z AUTO 06010G21KT 010V220 2 1/2SM HZ BKN007 22/M08 
A3008 $

2.5 miles of visibility, haze, and a broken layer of clouds at 700′ AGL when it’s 13% humidity. That seems rather unlikely. While that little ‘$’ does indicate possible need for sensor maintenance, I don’t think that was the major issue. Now, I wasn’t there, so I don’t know for sure. It is possible that the wind pushed under some humid air, lifting it enough to create a distinct cloud layer. More likely is that this was a dust storm and clouds – the sensor would just know there was obscuration, without knowing whether it was water vapor, dust or smoke. It is also possible that dust accumulated on the cloud sensor and that  the visibility was correct, but that the clouds were in error. The first human augmented METAR (a little over an hour later) helps confirm that dust was a likely contributor:

KCNO 021353Z 07015G29KT 030V090 3SM HZ BLDU CLR 24/M17 A3008
RMK AO2 PK WND 09029/1350 WSHFT 1310 SLP179 T02391167 $

No matter the cause, a pilot should see this and not be in a hurry to be anywhere around here flying. It gets a little stranger as we widen our view though. Brackett (POC) airport is just another 9 miles west from Ontario and little further north and closer to the mountains. At the same time as these dry gusty reports at CNO and ONT, POC is reporting

KPOC 021347Z 31003KT 10SM SKC 12/08 A3012

Still humid and light winds, but less than 10 miles away. That makes for quite a dividing line somewhere between ONT and POC. By the next report the air was drier but the winds hadn’t developed at POC. It took another seven hours before stronger winds and a shifted direction showed up at POC:

KPOC 022047Z 13015G20KT 10SM CLR 33/M15 A3011

Now, just keep going another 12 nm west of POC and you get to my home base of El Monte (EMT). I flew pattern work with a pre-solo student from 1500Z-1700Z with no problem. No turbulence and in fact we used runway 19, our normal runway (both ONT and CNO report wind gusts around 30kts out of the E/NE during this time). In fact, as I write this, EMT’s winds have never exceed 10kts and have continued to be out of the south and west.

KEMT 022346Z 21009KT 10SM CLR 33/02 A3001

Now, to show the local nature of these winds, head another airport west (18 nm WNW) of EMT to Burbank (BUR). Their winds have been mostly out of the south and light. The worst?:

KBUR 022053Z 17007G16KT 10SM CLR 33/M02 A3009

Keep going a little further west from BUR to Van Nuys (VNY) and now we get a nearly complete reversal of direction, despite VNY being only 7nm west of BUR:

KVNY 022051Z 02012G19KT 10SM CLR 34/M22 A3008

For those keeping score at home – that temperature/dewpoint spread equals 2% relative humidity. Is it little wonder that many parts of SoCal are burning? (and don’t forget the TFRs that often go with our local wildfires).

What should you take away from all of this? We have some crazy local weather. In these cases, the relationship of the airmass creating the Santa Ana and the terrain is key – ONT, CNO and VNY in these examples are downwind from two major mountain passes, while other airports may be more shielded from the mountains.When Santa Ana winds are forecast, proceed with caution – they frequently bring damaging winds and strong turbulence. Know that exact direction and intensity may vary from event to event and airport to airport and is hard to predict accurately.

2019 Update

As I was preparing a student for a checkride recently, we were faced with another forecast Santa Ana wind event. Other than TAFs, most of the aviation products lack the resolution to give good insight into the local variations of Santa Ana winds. I’ve recently found that the web site Windy can be helpful, though like any forecast, there’s limitations to its accuracy.

By default, Windy will center its map where it thinks you might be according to your computer IP address. It’s easy enough to drag the map around and zoom to show an area of interest. Using links to the right, you can show either steady wind velocities or gusts. You can also click in the map area to give a prediction for that point. A slider at the bottom allows you to change how far in the future you are looking. The “particle animation” (seen as ghostly moving arrows) is what really helps you see the odd circulations around the LA Basin, with strong NE winds below the Cajon pass, but more normal SW winds in much of the coastal basin. There are three different models that you can select from (Windy doesn’t generate the data, just visualizes what comes from common models). ECMWF seems to be the default (Windy is a Czech company and ECMWF is the European global model). So far, my experience says the NAM (a US regional model) may be the best selection for this particular use. The example below clearly shows the stronger winds near Ontario, Chino, and Corona.


Comment below if you have other questions or find this useful.

4 thoughts on “The Curious Case of Santa Ana

  1. Windy has an iPad app that I don’t use as often as perhaps I should, mostly because it’s hard to put down when I do. Now that Fall is upon us I’ll be paying more attention. 🙂

  2. Interestingly the ECMWF winds corresponded best with METAR reports during today’s Santa Ana event when compared to GFS and NAM models. Likely because ECMWF nonhydrostatic model accounts better for topography effects on weather systems

    • From what I’ve seen (small data set though), the ECMWF seems to underestimate intensities in these events and NAM is much closer (and seems to show the terrain impact much better). Right now is a perfect example. ECMWF shows nothing in SoCal over 20kts. NAM shows areas as high as high as 40kts, which is more in line with
      KDAG 280250Z AUTO 01025G34KT
      KEDW 280313Z AUTO 01028G31KT
      KWJF 280156Z 04020G26KT

Leave a Reply

Your email address will not be published. Required fields are marked *