Introducing RainCheck

Had a heavy thunderstorm yesterday evening after several weeks of dry weather. Most of this rain fell in a short amount of time, probably about 30-40 mins or so.

WF - .25"
Radar estimate - .50"
Raincheck - .46"

1 Like

I have two more days with light rain with similiar records.
WF 1.1mm
Manual gauge 0.7mm
WF 1.1mm
Manual gauge 0.6mm

1 Like

Heavy Storms passed through yesterday with a Tropical Storm on the way tomorrow.

Yesterday’s totals:
Weatherflow: 1.41"
Raincheck adjusted: 3.26"


Small rain yesterday.
WeatherFlow. 0.12
Percision. 0.03

1 Like

Nice storm last night but very brief.

WF - .25”
Radar estimate - .28-.32”
Raincheck - .33”

Overall very good performance with this one! Calibration continues to improve the results.


7/11/19 Rain event. Heavy, and very heavy, with 1 min of extreme.
WF- .92" ( Much improved, no longer under reporting heavy and above events.)
Actual- .89"
Radar estimate- 1.23"
Rain check- 1.19


Any idea when this will be available for European customers? My Sky and VP2 are mounted side by side , last Friday we had a high amount of rain and tthe difference between Sky and the VP2 is too big in my opinion.
VP2 registered 43,1 mm of rain, Sky only registered 19,4 mm’s , thats way way off!

1 Like

Last two days we have light rain and short rain shower
13.07 (light rain)

WF 1.5mm
Manual gauge 0.7mm

14.07 (drizzle, light rain, short shower)

WF 3.5mm
Manual gauge 3.1mm

1 Like

Attached spreadsheet gives various rainfall data over past two weeks

Rainfall.csv (409 Bytes)


Netatmo tipping bucket 0.42 inches
WF 0.22
WF w Rain check 0.43

1 Like

Yesterday we have light showers

WF 1.3mm
Manual gauge 1.1mm

1 Like

7/17/19 Rain events. Heavy,and very heavy.
WF- .25"
Actual- .24"
Rain check- .25"

1 Like

Ok…couple things here.

  1. As a meteorologist, I was honored to snag a field tester spot back a few years ago. Has been very fun testing this product out. Thanks for everything.
  2. I have been reading up on this rain issue for months, hoping it will get better, but the sad reality is that SKY cannot keep up with heavy rain here in FL, and I have to rely on raincheck to give me a rain amount closer to reality 8 hours later. How is this an appropriate solution for a realtime weather station?

An observation from east of Orlando last evening:
HEAVY storm (heaviest of the year) occurred 5:10pm through 7:25pm.
Manual gauge: 3.98"
Davis Nearby: 3.66"
WF (pre-rain check): 1.15"
WF (post- rain check): 1.89"

Is it just me or is this unacceptable considering the goal of the WF station is to report real time info? Tough to justify this station vs. a Davis for an extra few bucks at this point…

With all of that said, again I am a loyal field tester and love the concept here. But, in all reality, I paid for a station to do a certain thing and it unfortunately is not doing it. What is the short/long term fix here (because rain check, while helpful, is not accurate enough to warrant using the station for accurate rainfall info, in my opinion).

Thanks for any info you guys can provide!



Thunder storms through last night.
My WF was under reporting heavy rain before Rain Check. It didn’t take long for my WF to adjust but now it is over reporting rain by about as much as it was under reporting before. Unfortunately we haven’t had much rain since so I guess it will take longer to adjust downward.
Stratus .22"
WF .43"
Rain Check not in yet

1 Like

Ed1 your right on the money I was also very excited to receive a weatherflow station . I live in southern Indiana and mine works great in light to slightly moderate rain without much wind , but as soon as we get a microburst or a cell overhead the weatherflow severely under reports and 80% of the time I end up with someone else’s rainfall total a mile or 2 away after rain check . I have pretty much given up on rain check only time its accurate if you have a massive system that drops consistent rain over a wide spread area , at least in my case . I do think that we should have something that gives us real time data ( our Own Data ) . And I still look forward to checking it love the wind capabilities , I think weatherflow is way more accurate for light rainfall, around here in the summer a lot of rain evaporates before a tipping bucket measures it .
On the 16th we got some popcorn showers from the last hurricane and my Atlas was .51" 5-1 was .53" and wf raw was .36" and wf raincheck ended up at .38" 1/2mile west of me .67 1-1.5 miles southeast .37" 3/4 mile south ,47"

On the 17th which was basically the end of the popcorn showers the station 1-1.5miles south that yesterday rain check came closet too registered .34” , all the rest I mentioned above were within .07” - .09” wf raw was .05” ended up at .09”

I hope this is not the final solution but with so many variables I do not ever imagine getting a localized rainfall total



Hi Ed, RainCheck will help drive the Continuous Learning calibrations for your haptic sensor. There are many factors that influence accumulation accuracy including the intensity of the rainfall itself. Eventually, your SKY should achieve a better calibration for rain accumulation in general and these heavy precip events in particular. We can use your co-located gauges to refine calibrations for real-time reporting and work with MetStat on improving the reference product.
It looks like the heavy precip event was very localized. Your two reports are by far the highest when looking at CoCoRaHS data. Keep the reports coming!

Hi Tim,

Thanks for the reply. You’re exactly right - it was a very localized event centered in east-central Orange county (south of 408, east of 417). That 3.20" report was about 2 miles from me. I appreciate the reply and explanation, but I am still a bit confused as to the possible remedy for this situation. If this is being marketed as a weather station competitive with the big boys (i.e. Davis etc), shouldn’t there be a fix for this? I don’t think having to wait for rain check is an appropriate remedy for a weather station, especially in areas that regularly see extreme rain rates. I know you guys have been tweaking and improving as you go, which I admire having been with you guys from the beginning, but as a meteorologist, I just expected this to be a bit more refined since you’ve launched officially.

Regardless, the wind and temp is largely great which is awesome. I just wish this could be a full-package station, so I stop scrolling for Davis stations after every erroneous rainfall!

Any thoughts?


1 Like

Thanks, Ed. We are very thankful to have such dedicated field testers like you. Thanks for your early backing and continued support. We are actually just getting started with Smart Weather - don’t worry, we will be quite ‘competitive with the big boys’ soon!

Regarding your observations, is the Davis station to which you’re comparing available online?

The RainCheck value is actually designed to be an estimate of the “neighborhood” value - so it will not and should not necessarily match a co-located gauge during a highly variable rainfall event. As I’m sure you know, even two identical tipping buckets mounted relatively near each other can report significantly different accumulation values. RainCheck attempts to “smooth out” these values to give you a better estimate of the average rainfall across your entire property or neighborhood.

Remember that the haptic sensor is giving you more than a tipping bucket can: instant notification of rain start, fine grained relative intensity, total duration of rain, not to mention the lack of moving parts and minimal maintenance compared to traditional weather stations. Of course, you want accurate accumulation as well! The SKY should report daily accumulation to within 20% when well-sited.Your station seems to be well outside that spec at the moment. I just took a look and your station does not yet have a custom calibration (which are required to meet a variety of quality thresholds before they get applied). I’ve asked the team to manually review your data to see if yours can be updated. Please keep your reports coming.


One thing about raincheck, it does take a good number of rain events for the calibration to really kick in. Since raincheck started I’ve had maybe 15-20 rain events and it’s gotten significantly better with the heavy types of thunderstorms that it used to do poorly with.

A good example is a storm a month ago where I had 2+ inch per hour rain rates but my weatherflow maxed out at only .35” per hour which I could tell was way off. The storm total was .51 on weatherflow when in reality I ended up with almost 3” (2.75” per radar).

Now nearly a month later I’ve had several storms of similar intensity and have seen rain rates in the 1-2” range as reported on my WF station that match up very well with radar. My storm totals are within .10” of radar estimates whereas before a heavy storm they would be way off (but great with light to mod rain). Tonight we had a brief heavy cell move through, dropped .33” per WF and radar estimated .30-.35 for my area. I’m not sure how many calibrations have been done yet but I have seen significant improvements for my station. Perhaps as you get more rain events you will see similar improvements. My live WF data is pretty much spot on now no matter the rain type.


I hope so! Tim mentioned that my station was not being calibrated for some reason - hopefully things improve. I have full confidence in their ability! But you’re exactly right…my station is usually within 10% with light-mod rains, but struggles with anything heavy. I also improved the mounting of my station this week which I hope aids in increased accuracy!