Tropical Rain Versus Rainfall Estimates

[wx247]

Hi all. Hope you are having a great 4th of July weekend. I popped online for a few to check on the weather and I thought I would pose this question while there is some tropical downtime. While watching Alex come onshore there appeared to me to be a gross overestimation by the doppler radar sites of total rainfall from the storm... at least more than there is during a typical rainstorm. Is this usually the case with tropical systems?

I know that my rain gauge usually is pretty close to what doppler estimates around here.

Thanks for your help in answering my question!!!

[wxsouth]

Hi all. Hope you are having a great 4th of July weekend. I popped online for a few to check on the weather and I thought I would pose this question while there is some tropical downtime. While watching Alex come onshore there appeared to me to be a gross overestimation by the doppler radar sites of total rainfall from the storm... at least more than there is during a typical rainstorm. Is this usually the case with tropical systems?

I know that my rain gauge usually is pretty close to what doppler estimates around here.

Thanks for your help in answering my question!!!

Just to clarify, did you compare the doppler estimates to rain gauge amounts, and if so, were the gauge amounts much lower than the radar estimates?

[wx247]

Official rainfall totals from the NWS offices in Brownsville and Corpus both were much lower (in some cases 5-6") than estimates.

[thetruesms]

Interesting.

Perhaps I'm remembering incorrectly, but I thought radar tended to underestimate rainfall totals in tropical systems due to high rainfall efficiency and the tendency for tropical drop size distributions to skew towards smaller sizes than assumed in the 88-D's Z-R relationship?

[brunota2003]

If you are using the radars from those sites, it could be off because of ground clutter. Look at sites a good 30 miles or so from the radar and check it out then. Another thing is, did you verify the times? The radar totals may record 2 days, where official totals are day by day, thus giving the appearance of the radar being way off.

[wxsouth]

The radar totals may record 2 days, where official totals are day by day, thus giving the appearance of the radar being way off.

This is one possibility. Another is the use of the tropical z/r relationship. I'm guessing that these offices were using the tropical z/r to estimate rainfall. The keeps the radar from underestimating precip during tropical events. However...this relationship is really designed for rainfall estimates in the core of tropical cyclones. The droplet size in the outer rainbands differs from the inner core...and this may account for the overestimates.

[thetruesms]

The radar totals may record 2 days, where official totals are day by day, thus giving the appearance of the radar being way off.

This is one possibility. Another is the use of the tropical z/r relationship. I'm guessing that these offices were using the tropical z/r to estimate rainfall. The keeps the radar from underestimating precip during tropical events. However...this relationship is really designed for rainfall estimates in the core of tropical cyclones. The droplet size in the outer rainbands differs from the inner core...and this may account for the overestimates.

Yes, this could do it.

Also, are we talking about your own gauge, or ASOS-style gauges? Gauge error cannot be ignored, either. If it's a tipping bucket gauge, it can have some significant underestimates of its own in very heavy rainfall. It could have some underestimate in sustained high winds, as well.

[wx247]

I was comparing storm totals from the NWS radar versus the official storm totals released by those same offices. For example, there were some sections of Brownsville that the radar estimated having over a foot of rain (close to 14"). I never saw a report from the Brownsville area anywhere close to that number.

About the tropical z/r mode, I don't know when they have switched it on and when they haven't. I know that the amounts are UNDERestimated when they don't use this feature. That is what was throwing me for a loop.

[artist]

I have a question for the mets here. We have noticed up here just due west (barely) of west palm beach, lately, many times the radar is showing rain at our location yet not a drop is falling and vice versa. It used to be quite accurate, but when they had the major fix (it was down for awhile until it could be fixed from a major component going bad awhile back) is when I started noticing the discrepancy. Any ideas why that would be?