# How is Doppler radar used in rain prediction?

How is Doppler radar used in rain prediction?

• Do you have further context of what time of prediction they are talking? Doppler radar is great for predicting the next hour or so... as you can see what's aiming your way. But if they're saying Doppler helps predict hours down the road... it's not really all that useful. Strikes me as just a silly movie phrase, especially with the Doppler and Super Doppler phrase... just meant to sound silly and not very knowledgeable (Super Doppler is just a title some tv stations use for their Doppler!) Aug 8 '17 at 8:07
• Also towards the suggestion it's intended to be unknowledged or humorous: many forecasters, including the National Weather Service, don't vocalize forecasts of rain chances below 20% (you'll find the term "silent 10" in many forecast discussions by Googling). Plus typically forecasts are in 10% increments. So you generally won't see 5% basically anywhere except computer forecasts, at least in the US.. Aug 8 '17 at 8:10
• Does anyone else feel like the edits have lost the heart of the question? I wasn't sure what the user meant, but it seems there's a good chance we lost the core of it? Sep 4 '17 at 15:29

The easiest answer is that Safe Haven is a movie, and the writer made a factual error (which is not uncommon). I have not watched that movie, so I am unsure what type of situation they were dealing with. Doppler radars do not predict, but they observe.

What is possible, however, is extrapolation and the creation of inferences. For example, tornadoes are not directly predicted by Doppler Radar, but by making inferences from the data, a meteorologist may detect a tornado. In a similar sense, if the radar-derived 'storm total precipitation' has estimated an average 0.25 inches of rain from a squall line for the past 200 miles, one may extrapolate, or estimate, that the squall line will produce ~0.25 inches of rain. Could the squall line alter its path or intensity and drastically change the amount? Sure, but without additional information, including information not derived from a radar, that would be a difficult task.

Edit: Ok, I think I have a better idea of what the question was and how it can be answered. Doppler radar can measure wind speed relative to the radar site. So if can sense how fast a storm is approaching. Assuming the storm does not change it's speed, and given it's history, you can infer when it will arrive at the radar site. For example, if a storm is 50 miles away, and it moves at 25 miles per hour it will arrive in $50\text{ mi}\div 25\frac{\text{mi}}{\text{hr}}=2\text{ hr}$, provided the storm does not change.

First off, as others have pointed out, "Doppler" refers to the ability to determine the velocity of stuff (rain, snow, etc) towards/away from the radar.

Radar data are often assimilated to produce better numerical weather forecasts (i.e. using computer weather models). Typically the "main" variable of interest for this is the Doppler radial velocity (the velocity at which the rain/cloud/snow is moving towards/away from the radar) and NOT the radar reflectivity (which is related to the size/concentration of the rain/cloud/snow particles). Usually what you'll see on a TV weather station's radar display is something related to the reflectivity. The reasons why numerical weather forecast data assimilation mostly uses radial velocity are perhaps too complicated to get into here, but in short, the radial velocity provides information to the numerical weather model that's less complicated for the model to "use" than the reflectivity.

Doppler radial velocity from a radar is an essential tool for determining if a thunderstorm is rotating, and whether it might produce a tornado, or other damaging winds. Google terms like "tornado vortex signature" or "tornado velocity couplet" to see what this looks like. Generally for tornadoes the radar will show air moving towards the radar in close proximity to air moving away from the radar, indicating rotation.

Doppler radars are also used for more sophisticated research into cloud physics. For example, by pointing a radar vertically and observing the velocity at which rain/snow/hail falls, insight can be gained into the composition of these particles and the processes that grow them.

First rain -oversimplified. Rain occurs when water vapor in the air cools such that the air becomes saturated and the water vapor condenses as liquid water. It’s colder up high such that hot, vapor laden air raises, cools and releases rain. Lots of air and water moving about.

Doppler is the change in frequency of a wave as it bounces off a moving target. Radar is a wave that bounces off an object and detected back at the source. When radar bounce off water laden air moving upward (detected by the Doppler Effect) and cooling, it’s an indication that something rain like is occurring. Differing movements are indicative of differing parts of a storm.

Not long ago one or two planes full of people would crash at airport when microbursts of cool air would descend vertically and drive them into the ground. Since Doppler radar we can see these coming and avoid them.

• Doppler radar does not see air moving upward or downward. It only sees the radial component of the velocity of water particles moving toward or away from the radar antenna. The component of velocity orthogonal to the radial line is undetectable by radar. This is the zero isodop problem. Aug 9 '17 at 20:33
• My understanding is that such flow was turbulent and provided enough data for tracking. But I'm not hands on and will defer to a better knowledge.
– TomO
Aug 10 '17 at 15:47
• @DavidHammen but rising or falling air will have a component orthogonal to the radial lines from the radar! I mean, unless the radar is pointing directly up or down? ;-) Jul 9 '19 at 16:51
• @SemidiurnalSimon - Doppler detects range rate, the component of velocity along the line from the radar site to the target. The targets in this case, raindrops falling from clouds, is typically removed from the radar site by tens of kilometers. This means the vertical component of the falling rain is more or less undetectable, as is the horizontal component of velocity that is orthogonal to the line from the radar site to the clouds. Combining results from multiple doppler radar give a nice 2D view of the rain, but the vertical component remains more or less undetectable. Jul 9 '19 at 23:03
• @DavidHammen sorry, I got my language confused. But I also see your point that the radial component of far-away vertical movement is very small. Jul 9 '19 at 23:31

Some atmospheric models (namely the US NCEP HRRR) utilize radar as a source of initial conditions. I don't know the details of exactly how the radar data are assimilated. But Doppler data do provide the model a three-dimensional rendition of moisture and wind patterns on a scale that no other observing instrument is able to resolve, with our current technology.

Doppler radar is not used any differently to predict weather than older forms of radar were, satellite imagery is, or how physical observations that were then called or telegraphed ahead of an advancing system. The radar detects the presence of clouds with a water load that are likely to produce precipitation, or may already be doing so. That is compared to the direction the potential precipitation is moving and historical records of when similar conditions have occurred in the past, how likely is it to be repeated for a given location.

An observation is made, hey there is a cloud back forming there that seems to hold a lot of water. It is moving in this direction at this speed and the barometric pressure is doing this with a temperature like that. The last 10 times we had conditions that matched this, it rained 7 times, so, let's say there is a 70% chance it is going to rain.

Doppler radar is, in general terms, a more sensitive form of making these observations than older radar was, which was more sensitive and actuate than having physical observers a few hundred miles away calling and saying, hey, it's raining here. You will probably get rain in a few hours.

• Doppler radar is used quite differently to predict weather than older forms of radar. There's a reason the US National Weather Service upgraded their radar systems nationwide to NEXRAD in the 1990s. Doppler detects wind velocity (more specifically, range rate). Older radar systems did not. This added dimension significantly improved short term forecasts of severe weather. Aug 9 '17 at 20:20
• @DavidHammen I did not address how Doppler works. The OP asked how it is used to predict rain, not for an engineering explanation of how it functions of trivia like how until software AI was improved it could not tell the difference between a swarm of insects and a storm. I know about the vector differentials and side slip analysis which is analysed to signal potential wind sear and vortex formation, none of which is important to his question but is very important as to why millions of dollars were spent to upgrade to newer technology.
– dlb
Aug 9 '17 at 21:31
• The base answer is it sees conditions that indicate precipitation is possible. Those are compared to past similar conditions and a determination of the likelihood of a repeat is calculated and relayed. It is a vast improvement over look out a window and saying "Looks like rain".
– dlb
Aug 9 '17 at 21:33