Rain radar

[ronje]

There seems to be something strange going on with the Dampier/Pt Headland radars.

Each day it shows plenty of rain on the Pilbara coastline.

But if you look up the rainfall/river info, there's been no rain.

In addition the rivers/creeks water levels are flatline.

http://www.bom.gov.au/wa/flood/pilbara.shtml

Any ideas?

[Swoffa]

Pt Headland Automatic Gain Control (AGC) out of alignment.

Happens when they align it during the dry season with no humidity. As soon as a bit of humidity comes into the air you tend to get a lot of false echos.

Cheers

[ronje]

Thanks for that Swoffa.

Hard to believe that there's no air conditioning which allows stability irrespective of season (control of humidity).

They've probably recalibrated other radars in that area the same way.

You'd think that they'd have had a workaround by this stage.

[Swoffa]

Hey Ron,
no change to conditions inside the radar enclosure as they are climate controlled as you suggest. I was actually referring to atmospheric conditions that the radars are scanning out to 512km.

The calibration of the radars are done manually and adjusted to the conditions at that particular instance in time. Although the radars have an automatic gain control, any minute variation in atmospheric conditions after the calibration can have a significant effect on radar returns if the detection tolerances are set too low.

Similar if you manually set up the gains on your sounder for fresh water and then go through a more turbid or saltier section....increased noise on the screen and you would have to reduce the gain to suit. Unfortunately it might take a couple of months for the technicians to return on their normal maintenance schedule in these remote locations to adjust the gains or tolerance levels to compensate.

Regards
Warren

[Swoffa]

No surprise!
"The Pt Hedland radar is undergoing routine maintenance"

See how the picture looks after they have finished.

[ronje]

With you Warren.

Maybe not routine maintenance but that's better (with less liability) than saying "we got the settings wrong".

As a 512km image is a composite one, the 512km image for its neighbouring sites will probably carry the same apparent misinformation produced by the "dud" site/s.

With Pt Hedland off air, the other sites are still producing rubbish on other ranges so I'm guessing that there's a couple of other nearby radar sites that exhibit the same problem. eg Carnarvon, Learmonth and Dampier along with Broome which looks like it might need a touch-up as well.

Having 5 sites a bit wobbly suggests a servicing fly-by on the WA north-west coast would be of value (but not the same guy as last time).

[ronje]

Here's what the BOM Weather Connect Team says about it. Standard dismissive response.

Hi Ron

This appears to be caused by a strong inversion layer over the ocean. Radar beams ‘bounce’ between the ocean and the inversion layer “seeing” things that aren’t there.
Hope that helps!
Kind regards,

Weather Connect Team Leader
Community Outreach Program
National Forecast Services Group
Bureau of Meteorology

[ronje]

No, it doesn't help.

Inversion layers relate to temperature inversion with its attendant super-refraction.

That's not what's happening here.

The radars involved are C band radars (4-8 ghz) with the ability to see smaller targets (shorter wavelengths) and THAT'S why they are displaying returns.

These returns visible now would have not been visible to the S band radars (2-4 ghz) with their longer wavelength (nothing that a simple adjustment of the sensitivity couldn't fix).

The inversion layer must have been in place for the last 2 months day and night ('cos the radar outputs don't change) and exist over the sea AND land. Temp inversion doesn't last 2 months without change day and night

Simply turn the sensitivity threshold (or the AGC) down. These radars have apparent sensitivity of -114dbm. The smaller wavelengths allow more eirp for similar sized dishes as S band dishes. So with higher eirp, increased sensitivity and shorter wavelength frequencies involved the ingredients for adjustment error are there.

regards
Ron