progress audio camera moved down camera moved left camera moved right camera moved up free exploration mode guided tour mode hotspot Use the w, a, s, and d keys to move around the 360 degree image. Press the tab key to jump to interactive markers and hotspots. 360 degree image interaction label marker items next pause audio pause video play audio play video previous visited of %count% of %total% step %count% of %total% item step %total% items %total% item progress video video volume %count% of %total% items visited %count% of %total% item visited %count% of %total% visited Accessible text Keyboard shortcuts Zoom to fit active Alt Check to include Clear and return to menu Close Closed captions. Press the spacebar to toggle captions on and off. Toggle captions Continue Continue Ctrl Enable keyboard shortcuts Enter Exit Filter FINISH finish Definition definition Glossary Terms hour hours Action Keyboard Shortcuts List keyboard shortcuts Mute / unmute Next slide Play / pause Previous slide Replay slide Shortcut Submit slide Toggle accessible text Toggle closed captions Toggle zoom to fit Menu Please rotate your device minute minutes Mute NEXT next Next slide Notes Notes Bio More info Captions pause Pause play Play Send an email PREV previous Previous slide Question List Replay replay Resources resources Restart Resume locked search Search Sidebar Toggle slide progress Start Course Start Course visited Search... Search in: Search Results second seconds Settings Shift Skip navigation. Press enter to return to the slide. Slide Text Space SUBMIT ALL submit all SUBMIT submit Submit Unmute volume <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Span Text="Welcome to Principles of Radar! My name is Katy Christian and I’ll be your guide through today’s lesson over Rainfall Estimation Errors." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="If you are completing this course for credit, please go ahead and review this interaction on how to complete this course within the NWS Learning Center. Once you’re done, click “Next” to move on. " /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Span Text="Just for some context, here is the roadmap for the entire Radar Principles section of RAC - right now, you are here so let’s keep moving forward! "> <Style FontFamily="+minor" FontSize="9" ForegroundColor="#444444" LinkColor="#444444" /> </Span> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="By the end of this lesson, you should be able to name the most common errors in rainfall estimation, explain what causes those errors, and lastly, be able to identify which errors lead to either an under- or over-estimation in rainfall." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="Remember from the previous modules, we talked about two different algorithms for estimating rainfall. The first estimates rainfall using legacy base data and is referred to as the Legacy Precipitation Processing Subsystem, or Legacy PPS. The second algorithm uses BOTH legacy and dual-pol base data to estimate rainfall data and is referred to as the QPE algorithm. In this module, we are going to talk about several common errors that impact rainfall estimates in both of these algorithms.
" /> </Block> <Block> <Style /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="Allright, below is a roadmap of the most common sources of rainfall estimation errors. To learn more about each one, simply click on the corresponding icon. Once you are done exploring each topic, click next to continue." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Span Text="You probably already have a general idea of what clutter is, but how does it happen? Clutter errors can occur when the clutter filtering algorithm is turned off or when clutter remains after filtering. Remember from an earlier module that the clutter algorithm works only if the clutter echoes are NOT moving. So, clutter will not be filtered out if any of the sources have some movement associated with them. So, what are these? Common moving sources of clutter include wind farms, moving cars, insects, birds, airplanes, and Superman if real.
" /> </Block> <Block> <Span Text="
" /> </Block> <Block> <Span Text="Take a look at the image below. Multiple storms are ongoing at this time as well as some clutter that did not get filtered out. Can you guess where it is? Click on the image where you think it’s occurring and I’ll let you know once you’ve got it. If you guessed this area, you are correct! These stationary high, noisy reflectivities are actually the result of a wind farm here! You can confirm this by toggling over to correlation coefficient and making sure you have low CC values, which we do indeed see we have. You should also see low velocity and spectrum width values.
" /> </Block> <Block> <Span Text="
" /> </Block> <Block> <Span Text="So, how does clutter impact precipitation estimates? If you’re not careful, clutter may be incorrectly interpreted as precipitation echoes, especially when they occur in the midst of ongoing storms as we see in the figure here. Therefore, it's important for you as a forecaster to be aware of what various types of clutter look like and know what types of clutter frequent your CWA." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="Remember that the relationships that convert reflectivity to rainfall rate, or Z-R relationships, are empirically derived and fixed. If you are using the Legacy PPS for rainfall estimates, here are the 5 different Z-R Relationships and the best environments to use them in. Because it is up to the forecaster to choose the appropriate Z-R relationship for a given event, errors in precipitation estimates may creep in if the forecaster is not vigilant in choosing the Z-R relationship that most closely mirrors the precipitation event type. These errors can either be drastic under- or over-estimation errors.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="An example of this is shown in the below images from a study in 2009 over historic flash flooding that occurred in central Kansas in August of 2003. The two images show precipitation estimates for the same event but from using two different Z-R relationships: the convective Z-R relationship on the left and the tropical Z-R relationship on the right. As you can see, the difference in precipitation estimates from Z-R relationships alone is quite drastic - as much as 3 inches in difference! For this historic event, the convective Z-R relationship severely underestimated rainfall amounts while the tropical relationship more closely mirrored local gauge amounts. 
" /> </Block> <Block> <Style /> <Span Text="So question time: Let’s say you work at the Fort Worth office and it’s September. Your office recently had lingering remnants from a hurricane move through so you switched to the “Tropical” relationship. It’s been about a week now and you expect widespread light rains from an incoming trough. Which Z-R Relationship should you switch to? Click below to select your answer! 
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="BEAM BLOCKAGE: Another source of precipitation error is when the radar beam is either partially or completely blocked by geographic features or buildings. In the image shown here, can you guess where beam blockage is occurring? Click on the image where you think it’s occurring and I’ll let you know once you’ve got it.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="If you guessed this region, you are correct! Several trees to the southwest of the radar partially block the radar beam. While beam blockage can occur anywhere, it is especially prevalent in the western United States where large mountains often completely block the radar beam.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="In cases where the radar beam is partially blocked by terrain, any precipitation that is detected beyond the blockage will be underestimated as the hydrometeors will be sampled by a reduced power beam and will have a diminished return.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="Now, not all is lost in cases of beam blockage. Remember from the QPE Algorithm module that there IS a rain-rate relationship in the DP QPE algorithm that can estimate precip in areas of beam blockage as long as those areas are strictly beneath the melting layer. Do you remember which rain rate equation in DP that was? Click which one you think that was." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="Over time and use, a radar may fail to meet certain calibration standards. If you notice that precipitation estimates from your radar routinely do not match with gauge observations and seem unreasonably high or low, you may want to check to see if it is correctly calibrated. You can do this a couple of ways. To check calibration using reflectivity, you can find this on the RPG Control Status page under the Environmental Tab. The highlighted green box displays calibration values for both the horizontal and vertical channels. For calibration using reflectivity, these values should have a bias less than +/-1 dB. So for the example shown here, the reflectivity calibration values are in good shape!
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="For calibration using differential reflectivity, you can either check the data quality dashboard page or the shift change checklist. For differential reflectivity, calibration values should have a bias less than +/- 0.2 dB so in this example from the data quality dashboard page, our ZDR bias is 0.28 dB which is outside our calibration standard of 0.2 dB hence the orange color indicating something is off.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="Here is another example but from the shift change checklist where the calibration values for differential reflectivity can be found here - do you think this radar is correctly calibrated? Click to select your answer. 
" /> </Block> <Block> <Style /> <Span Text="That’s right, this radar is definitely not correctly calibrated according to ZDR values and it should be further investigated by the electronic maintenance staff. Calibration errors like this can result in large errors in precipitation estimates, whether they be under or overestimation errors. 
" /> </Block> <Block> <Style /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="MELTING LAYER:The melting layer, or “bright band”, is a layer within the atmosphere where hydrometeors transition from solid to liquid. To understand why the melting layer creates a precipitation estimate issue, remember from earlier that radars assume Rayleigh scattering where the particles detected are assumed to be much smaller than the radiation emitted. However, in the layer, Rayleigh scattering is violated because very large hydrometeors exist. This acts to greatly enhance reflectivities and results in a large overestimation of precipitation, sometimes up to five times more than what they should be. 
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="In the example here, can you guess where the bright band is? Click the general region where you think it is and I’ll let you know when you’ve got it.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="If you guessed this area, you are correct! In this region, you can see a band of enhanced reflectivities surrounding the radar and if we toggle over to correlation coefficient, you can see CC values around 0.9, indicating mixed phase precipitation within the melting layer. One way to prevent incorrectly identifying the bright band as a measure of surface rainfall is to know at what level the 0 degree isotherm exists." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Span Text="The effects of range are a fundamental challenge when using radar to estimate rainfall. One of the biggest estimation errors is when the radar beam overshoots precipitation, resulting in a gross underestimation or the radar not even sensing the precipitation. Overshooting is particularly a problem during the cool season when low cloud bases with shallow precipitation are common. For most radars, overshooting becomes a major issue at long ranges, so anything > 60 nautical miles.
" /> </Block> <Block> <Span Text="
" /> </Block> <Block> <Span Text="Another artifact of range is the reverse where instead of overshooting the precipitation, the radar beam detects rain aloft but due to evaporation below, all the rain is evaporated before it reaches the ground so the radar is grossly overestimating precipitation. This is often the case in dry desert environments in the West where deep, dry sub-cloud layers evaporate any falling precipitation. So, just be aware that the further you are from the radar, the greater the chance the radar could be overshooting precipitation or indicating rain even though it never makes it to the surface.
" /> </Block> <Block /> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="What is attenuation? Attenuation simply refers to a reduction in power transmitted and returned from the radar. There are a few things that can cause attenuation including heavy rain, large hail cores, and when the radar is covered by a thin film of water, referred to as a wet radome.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="In the case of heavy rain, when a radar beam encounters large concentrations of hydrometeors, the returned power signal is weakened and the radar has a hard time seeing very far into the heavy rain. As such, attenuation due to heavy rainfall results in an underestimation of precipitation. In the below example, can you guess where attenuation is occurring? Click on the image where you think it’s occurring and I’ll let you know once you’ve got it. 
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="If you guessed to the northwest of the radar, you are correct! Both heavy rain and a hail core are causing a reduction in power down radial resulting in an underestimation of the actual amount of precipitation that is falling in this area. If this persists over several radar scans, precipitation estimates here will suffer from underestimation.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="As for the case of a wet radome, under certain circumstances a thin film of water can coat the radar and act to reduce the power transmitted and returned from the radar. This results in an underestimation of rainfall over the entire radar field. In the below example a squall line is approaching the radar. In the next image, you can see a visible reduction in the magnitude of reflectivities across the entire radar field of view." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="PARTIAL/NON-UNIFORM BEAM FILLING:Alright, let’s talk about what causes both partial and non-uniform beam filling. Partial-beam filling occurs at long ranges from the radar when precipitation only fills a small fraction of the beam as seen in this image. Since the radar equation averages the reflectivity over the pulse volume, the depicted reflectivity will yield lower values than what is actually occurring in that fraction of the beam as shown here. However, since the radar beam widens at increasing distances, this underestimation will be spread over a much wider area and will result in an areal overestimation of precipitation as shown.
" /> </Block> <Block> <Style /> <Span Text="
" /> </Block> <Block> <Style /> <Span Text="Non-uniform beam filling occurs when a gradient in precipitation type exists across the beam which is more common at increasing distances due to a larger beam width. The change in differential phase due to different hydrometeor types largely affects correlation coefficient and results in an underestimation of CC values down radial of where this gradient exists. An example of this is shown here - notice the radials of dampened CC values to the southwest of the radar. When CC is underestimated down radial, active precipitation areas down that same radial may not be converted to rainfall. 
" /> </Block> <Block> <Style /> <Span Text="Both partial and non-uniform beam filling become a more common issue at increasing distances from the radar where the beam width widens." /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Style /> <Span Text="Let’s now recap very quickly what we’ve learned so far: there are two types of rainfall estimation algorithms - the Legacy PPS is based solely on Legacy base data whereas QPE is based on both Legacy and Dual-Pol base data. Both these rainfall estimation algorithms suffer from numerous errors that we discussed and are listed below. Allright, you are now finished with the lesson portion of this module. Whenever you’re ready, please click next to move on to take the quiz. " /> </Block> </Content> </Document> <Document xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <Content> <Block> <Span Text="Congratulations on passing the quiz and completing this module! You have lived to see another module, of which you can see on this image here what comes next. Once you’re finished, click the “Exit” button to close this lesson. "> <Style FontFamily="+minor" FontSize="9" ForegroundColor="#444444" LinkColor="#444444" /> </Span> </Block> </Content> </Document>