pws Weather Station Experiments and Brand Comparisons

[kinzyjr]

Foreword

Performing a quick search on this site will yield a significant amount of threads, recommendations, and opinions on the various brands of weather stations in regard to their accuracy and feature sets.  A quick list includes:

• https://www.palmtalk.org/forum/topic/76321-my-new-weather-station
• https://www.palmtalk.org/forum/index.php?/topic/66495-best-home-weather-station/
• https://www.palmtalk.org/forum/index.php?/topic/66232-outdoor-thermometer-recommendations/
• https://www.palmtalk.org/forum/index.php?/topic/59865-weather-stations/
• https://www.palmtalk.org/forum/index.php?/topic/55505-do-you-have-a-weather-station/
• https://www.palmtalk.org/forum/index.php?/topic/39264-rapid-fire-weather-station-model-recommendations/
• https://www.palmtalk.org/forum/index.php?/topic/34877-weather-stations-digital-indooroutdoor-thermometers/
• https://www.palmtalk.org/forum/index.php?/topic/21182-best-weather-stations/
• https://www.palmtalk.org/forum/index.php?/topic/16414-weather-stations-for-the-home/
• https://www.palmtalk.org/forum/index.php?/topic/9319-where-do-you-measure-your-temperature/
• https://www.palmtalk.org/forum/topic/58989-my-first-weather-station/
• https://www.palmtalk.org/forum/topic/26652-home-weather-station/
• https://www.palmtalk.org/forum/topic/17019-wireless-weather-station/
• https://www.palmtalk.org/forum/topic/72264-datalogger-recommendations-for-microclimates/

Most of us record our own garden temperature as it can vary from the local airport due to microclimate factors that are not conducive for official readings.  This thread attempts to capture the experiments done to compare some sensors in my personal collection against some of the better personal weather station (PWS) brands, as well as benchmarking them against the various sensors used by the National Weather Service (NWS).

PWS Brands in operation locally:

• AcuRite: These are available as a complete weather station, individual sensors with various Min/Max recorders, and red liquid thermometers
• AmbientWeather: The 2000, 2902, and 5000 models are frequently used and marketed as a great mid-grade PWS
• Avantek: Thus far, I haven't seen anyone else use this brand.  Mine was a gift.
• BloomSky: Very few of these in my area, but the Sky2 does capture temperature
• Davis: The most commonly used models are the Davis Vantage Pro 2 and the Davis Vantage Vue. 
  • My Bias: A Davis Vantage Pro 2 Plus with a fan-aspirated radiation shield is the gold standard for a PWS.
  • The station in question: https://tinyurl.com/yrpyn9dr
• LaCrosse: There are a few of these around and Lowe's carries some that are similar to the AcuRite sensors.
• Logia: The typical setup in operation is the Logia 5-in-1
• Netamo: There are a few of these around, but they are not common
• Onset HOBO: A Davis Vantage Pro 2 looks expensive until you look at some of these.
  • Website: https://tinyurl.com/2ck4vufs
  • ABS radiation shield: https://tinyurl.com/2p8y479y
• Oregon Scientific: The most common model in operation in my region is the WMR2000
• RainWise: Uncommon here
• Tycon Systems: Uncommon here
• WeatherFlow: There are few spread throughout my region

The thread will contain several experiments conducted over time that fit together as a whole to provide as much salient information as possible in regard to the value you get for your bucks.  My threads typically include file downloads that have some utility for understanding the work behind the content.  This thread contains a spreadsheet of Weather Underground station information for the Lakeland area, with the Davis Vantage Pro 2 units highlighted.  Some of the stations are no longer in operation.  This sheet is useful for comparisons by brand.

202302132345_LakelandWeatherUndergroundStations.xlsx

[kinzyjr]

Experiment #1: How close (or not) is a low or mid-grade station to a Davis Vantage Pro 2?

Setup:

02/06/2023: On this night, the temperature was only slightly cooler than average.  The idea was to perform this check on a night closer to normal to see if the numbers get further apart as the temperature deviates further from average.  An AcuRite Min/Max collector/sensor combo was set outside at sunset and given ~90 minutes to acclimate.  The temperature matched an AmbientWeather 2902D with the latest firmware installed.  The AcuRite was placed within 50 yards of a Davis Vantage Pro 2 at a height of 5 feed with the Min/Max receiver directly underneath it on the ground at 8PM.  At ~9:50PM, the sensors were compared.

02/13/2023: The goal was to check the difference on a clear night, especially one with cooler than normal temperatures expected.  The chilly night did not necessarily need to near freeze territory, so long as it captured the difference between devices.  The first comparison was done with an AcuRite Min/Max and an Ambient Weather 2902D with the latest firmware installed.  Since the AcuRite sensor has no radiation shield, the comparison must be done after sundown.  When the sensor was placed outside, it was given two hours to acclimate to the outside.  The temperature difference was 1^oF at 8:30pm.  The sensor was then transported to a location within 50 yards of a Davis Vantage Pro 2 and placed at a height of 6 feet, with the Min/Max receiver directly underneath it on the ground.  Everything was set at 8:50PM and left to sit until ~10:20 (90 minutes).

Results:

The AcuRite and Davis stations were within 50 yards of each other with the same elevation and same landmass.  The AcuRite consistently records the lower temperature by ~3F+/- 0.5.  The airport readings are included, but this area is too far away for a direct comparison.

02/06/2023

02/13/2023

[Sabal_Louisiana]

It has been my experience that Acu-Rite stations trend toward heat bias during warm sunny days despite having an aspirated fan and under read, as you have shown, cold clear mornings. Also, they sometimes have issues with their rain gage. Perhaps the more expensive Atlas series are of better quality.

I've had two Ambient stations and none have lasted more than three years but I have two co-workers that have been running theirs for several years. The temp and precip readings are usually accurate but I find that the most popular 2902 model under reads solar radiation and UV in the summer.

The most advanced Ambient is the ultrasonic WS5000, which costs over twice that of the WS2902.

I have a Logia 7 in 1 and the temperature readings seem to be spot on, even on hot days, and perform a little better than Ambient. Precip matched against manual rain gage is typically within a few .01 inches if not identical. Oddly, when I first had it, these readings were off a bit but after a couple of weeks adjusted rather well. UV and solar radiation seems to be pretty accurate, too. The only problem is after 9 months of use, it could no longer connect to wifi and I am awaiting a replacement.

Without spending a fortune, Davis is still the best. But even they are more much more expensive than the other common models and you cannot readily upgrade a Vantage Vue with the components of a Pro 2 Plus, like solar radiation and UV. Also, the Pro 2 is rather bulky in size so if space constraints are an issue, it is a challenge to place it.

WeatherFlow has a Tempest  system ($330) with no moving parts, the entire unit looks like a microphone. It utilizes a haptic rain sensor. I've been curious about trying it out.

[flplantguy]

19 hours ago, Sabal_Louisiana said:

WeatherFlow has a Tempest  system ($330) with no moving parts, the entire unit looks like a microphone. It utilizes a haptic rain sensor. I've been curious about trying it out.

I was looking at this and their network and it looks good, but its new tech somewhat so im wondering how far that money goes. If it works well ill get one for my yard as my sensors are cheap and innaccurate as far as i can tell.

[kinzyjr]

@Sabal_Louisiana + @flplantguy

This is probably the PWS you are referring to: https://shop.weatherflow.com/products/tempest

Since there were some cold spells in January 2022, December 2022, and January 2023 and there are a few Tempest stations close to a Davis Vantage Pro 2 in my area, I moved this comparison to the top of the list to give you some data to consider.  The December 2022 cold snap was mostly advective in nature, so there was less variation in the air temperature due to microclimates.  Here are the stations I used from WeatherUnderground with approximate distances in feet from the DVP2:

The figure most PalmTalker's are concerned with is the absolute minimum:

• The minimums in January 2022 range from 28.9 to 31.5, with the airport recording 30F. 
  • The damage report from this freeze is here: January 2022 Florida Freeze Report
• The minimums for December 2022 range from 29.8 to 30.2. 
  • The official NOAA low in the station download is 29F, but I did capture a recording of 28F at the airport from NWS.  (see screenshot at the bottom)
• The minimums for January 2023 ranged from 32.0-33.6, with the airport recording 34F.

One thing that stands out to me statistically is that the airport seems to edge out these stations in regard to the average low recorded over a month.  When the absolute low is recorded, the stations tend to record higher numbers.  Anecdotally, this seems to happen with other brands as well.

The High temperature tends to generate more discrepancy, but not so much that any of these stations are ineffective. The average temperatures over a month tend to have a wider tolerance as well, but in both cases, we're looking at ~1-4 degrees.  Deviation from the airport cannot necessarily be an indictment of these stations' accuracy since they are 8 miles away and on rolling hills rather than flat land.  This data is more for comparison.  The records and their rounded equivalents are included in the attached spreadsheet if anyone would like to review it.

There are more of these stations in the immediate area, but a few of them are mounted on the roof of a building, making their records unreliable.

Screenshot of NWS Lakeland Airport (KLAL) on December 24th/25th, 2022 (originally posted in Florida Christmas Freeze 2022 thread):

202302162100_DavisTempestCompare.xlsx

[flplantguy]

Does a radiation shield effect nighttime lows?  A sensor exposed to the sky would radiate like a car roof, but under cover wouldn't perform the same, right?  I doubt that is part of the difference with these, but it was a thought about my own sensors at christmas and both january events.  Especially as the frost on jan. 16 was only JUST starting to form on the palmetto leaves at head height when the sun struck the leaves.  Coldest sensor in the open was 31 so it makes sense, but if it was not done correctly the second reading was 32.4 and still could have frost, right? So it may even be a moot point idk.  Ive been told to obsess over my health the same way lol.

[kinzyjr]

1 hour ago, flplantguy said:

Does a radiation shield effect nighttime lows?  A sensor exposed to the sky would radiate like a car roof, but under cover wouldn't perform the same, right?  I doubt that is part of the difference with these, but it was a thought about my own sensors at christmas and both january events.  Especially as the frost on jan. 16 was only JUST starting to form on the palmetto leaves at head height when the sun struck the leaves.  Coldest sensor in the open was 31 so it makes sense, but if it was not done correctly the second reading was 32.4 and still could have frost, right? So it may even be a moot point idk.  Ive been told to obsess over my health the same way lol.

Does a radiation shield affect nighttime lows?  A sensor exposed to the sky would radiate like a car roof, but under cover wouldn't perform the same, right? 

It shouldn't.  If the shield is touching the sensor, the sensor could mistakenly read the temperature of the shield rather than the air temperature. 

As for reading a temperature under cover, you might notice a temperature difference if you record a temperature under tree canopy, but if the air circulation is good around a temperature sensor, it shouldn't matter.  If you have enough equipment, it might be worth running a sensor 'naked' vs. on with a shield.  I've had to do maintenance on a few at dark and the temperature didn't seem like it changed when I removed the shield.

Ideally, for receiving accurate "official" air temperature and wind readings, you'll want the following characteristics in a weather station:

• Separate components for measuring temperature and wind.
  • This allows you to put these components at the differing heights recommended for temperature/humidity/dew point and wind speed/direction
• A large radiation shield or shelter that is fan-aspirated
  • The large shield ensures that no part of the shield makes contact with the sensor.
  • The fan aspiration guarantees good air circulation so you are not measuring temperature in an "oven" where heat builds up and corrupts the reading
  • The radiation shield should be made out of white, ABS plastic. 
  • The plastic itself should be white, not just painted white.
  • If I were doing a "build it yourself" rig, I'd grab the Type B ABS shields shown below, get at least 8 floors, and fan-aspirate it.
• A few resources from weather.gov are attached for more complete guidance and examples of do's and dont's in regard to siting.
  • Few of us living in urban or suburban areas have the space to fulfill the requirements for an official reading.  My site is no exception.

Coldest sensor in the open was 31 so it makes sense, but if it was not done correctly the second reading was 32.4 and still could have frost, right?

It's very common during radiational cold events to have differing readings throughout your property.  @RedRabbit documented quite a temperature spread on his property this winter.  In my case, I typically put out 5 sensors during cold events, but skimped this year since there were other pressing issues that demanded my attention.  During the January 2022 cold snap, I recorded 31F as the lowest and 34F as the highest temperature.  That's consistent with the DVP2 on the other side of the hill, so it's within acceptable parameters.

You'll notice that frost can form well above freezing in open areas.  In January 2022, I observed frost at 39F.  Previous to that, I had observed frost as high as 37F.

A good portion of my back yard is covered with live oak canopy, so I have limited areas where I can measure temperature and consider it a reasonable estimate of the air temperature.  Anything measured there is "garden temperature."

Sample Radiation Shields from AliExpress (only for demo - I haven't bought these myself in the past😞

 

CWOP-Siting.pdf 202302171330_COOP_ProperSiting.docx

[kinzyjr]

Experiment #2: Benchmark personal equipment against an airport sensor

• How close is an AcuRite electronic sensor to a Nimbus Min/Max Temperature Sensor (MMTS) at an airport?
• How close is an AcuRite red liquid thermometer to a Nimbus Min/Max Temperature Sensor (MMTS) at an airport?

Setup: Since neither the electronic sensor nor the red liquid thermometer by AcuRite are equipped with a radiation shield, readings must be done at night.  Additionally, there are few places where one can perform this experiment within a short enough distance to get reliable results without getting arrested.  The Nimbus sensor at KLAL is relatively close to the parking lot, some open grass areas, and two hotels.  At night, the NWS typically takes records at XX:15, XX:35, XX:50, and XX55 at night at KLAL.  The goal is to arrive at least 60 minutes before readings start, while keeping the equipment as close to ambient air temperature on the way there by riding over bundled up with all four windows open.  After the sensor or thermometer is placed, wait 60 minutes and begin documenting the temperature at the specified times.  The equipment used is shown below the results.

Results:

The results above show that AcuRite equipment tends to record temperatures below what is recorded with Nimbus equipment used at small airports.  For this test, the sensors were hanged from a tree at a height of approximately 5 feet out in the parking lot. While this may not be congruent with official record siting requirements, the lower readings were expected and the goal was for the bias to be in a positive direction.  The site for the airport sensor is between the parking lot and a large asphalt area as well, so this is not entirely different than the surface near the airport sensor.

If you use these sensors and recorded a temperature that was inordinately low, you could safely add a degree or two when comparing what you get to the airport nearest to you to determine if you're in a cold hole on that particular night.  I know @Walt has used these, and I've used them myself at points.

ASOS stations may be a different story, but testing if first-hand would require a bit of travel to perform a direct comparison.

The AcuRite Red Liquid ( available at Big Blue Box ) :

The AcuRite Electronic Sensor ( available at Big Blue Box ) :

KLAL Sensor Location and Equipment Used (USC00084802) :

KLAL Photo of Sensor from the Parking Lot (Daytime):

[RedRabbit]

So, you’ve effectively audited the airport sensors and proven definitively they’re reading too high on cold nights? 
 

Wow, way to go @kinzyjr!

[JLM]

You should present this data to the NWS. Its possible that there may be something that they can do to fix this high bias. Wouldnt get my hopes up though.

[flplantguy]

They acknowledged it a bit after the mistake in forecasting in i think 96.  They used airport data for all the forecasts and farmers were not prepared.  This is so detailed though im sure its new to them but it also matches what changed at brooksville recently.  It seemed to show warmer than the area sensors in jan 2022 than most WU stations by nearly 3 degrees when it used to drop like a rock.  Part airport schew and part home sensor error perhaps adding to a few degrees discrepency?

[kinzyjr]

@RedRabbit + @JLM + @flplantguy

It could go one of a couple of ways:

• Given that Experiment #1 shows the AcuRite was ~3F lower than the DVP2 and ~1F-2F lower than Nimbus at a given time, either:
  • The AcuRite is more accurate, and the other stations are measuring too warm
  • The DVP2 and Nimbus are more accurate, and the AcuRite is measuring too cold
    • By corollary, the WeatherFlow Tempest was really close to the Davis.  The same statement would stand for that brand as well.
  • All are inaccurate, resulting in only a best guess within a +/-3F tolerance.
    • Since the tolerance on most low and mid-grade stations is +/-2F and the tolerance on premium brands like Davis is +/-1F, you could make a good case for this.

There is another phenomenon that I've looked at, where a few AmbientWeather stations I've checked seem to initially have their temperature recordings drop off a cliff for a few hours when the sun goes down, but then arrive fairly close to the DVP2 and airport in regard to the nightly minimum.  The difference can be a 4F-7F difference compared to the DVP2 and/or KLAL from the observations I've made.  Since there is some high temperature elevation during the daytime, my guess is this is due to influence from the heat shield.  This will be a good test item when I figure out a way to test it.

It's been interesting.

[kinzyjr]

Experiment #3: How consistent is a Davis Vantage Pro 2?

• Consistent highs? lows?

Setup:

In order to test a station against itself, you would need two stations of the same brand very close to each other.  It is preferable for at least one of the stations to be hardwired to reduce the chance of wireless signals conflicting with each other.  In order to find two DVP2s within 300 feet of each other, with at least one being hardwired, these are just what we need:

• https://www.wunderground.com/dashboard/pws/KFLLAKEL51/
• https://www.wunderground.com/dashboard/pws/KFLLAKEL251/

We can investigate records from some of our recent cold fronts on these two stations and gauge how far apart we could expect the highs and lows to be apart on two units under the same conditions.  The cold snaps in February 2021, January 2022, December 2022, and January 2023 were used to observe the differences.

Results:

The billed tolerance on a DVP2 is +/-1F.  On average, this would mean that the range should be less than or equal to a maximum of 2F (e.g. a station reads 1F high, the other 1F low = 2F difference).  The high temperature tends to have more variance than the low temperature.  For our purposes of growing palms, this is fine.  Both averages are well within 2F variance, hence this would meet its specifications.  The lows are incredibly tight for a PWS.  There can still be some considerable variance between two very accurate stations over the course of a month as you can see in the max_delta and min_delta columns.  A positive number means the station in the first bulletpoint above was milder than the second, while a negative number indicates the converse.  The day-by-day numbers for each station are included in the attached Excel file.

NOTE: I had concerns that my 28.2F recording during the Jan. 2018 Freeze was suspect considering the airport recorded 25F and the freeze was advective.  My weather station was an Avantek at the time.  The DVP2 station in Experiment #1 is reasonably close by, so I decided to post a quick double-check I did on it (also note the wind speeds recorded at ~5ft high):

202302212145_DVP2_Compare.xlsx

[kinzyjr]

Experiment #4: How does and Ambient Weather 2902 compare to a Davis Vantage Pro 2?

Setup:

Ideally, we could compare two stations on the same property.  The next best thing is to find two stations that are relatively close together.  Lakeland presents challenges to comparing stations, even if they are in close proximity, due to changes in elevation, proximity to large lakes, and a narrow heat island that dissipates quickly outside of the urban core.  The ideal pair of stations should be located outside of the urban core, away from any large lakes, and in a stretch of land that is relatively uniform in elevation.  The northwest area of town fits this bill well.  As such, the two stations selected were:

• Davis Vantage Pro 2 (KFLLAKEL57): https://www.wunderground.com/dashboard/pws/KFLLAKEL57/table/2021-04-1/2021-04-1/daily
• Ambient Weather 2902 (KFLLAKEL167): https://www.wunderground.com/dashboard/pws/KFLLAKEL167/table/2023-06-23/2023-06-23/daily

The straight line distance between the two stations is approximately 950 ft.

Both are in a suburban setting in the northwest quadrant of the city and not near a body of water large enough to significantly modify the climate.  In spite of the stations' reported difference in elevation, the topographical map shows that both are very close in elevation (~200ft above sea level) .

Both stations have an overlap in records from August 2018 through April 2021.  Records were collected and compiled by the month, with NWS records added for comparison.  If there was a significant deviation (>= +/-5F), then the cell containing the temperature delta was set to change colors for easy flagging.  Please note that while NWS records are included, this area is at a higher elevation than the airport and closer to the US-98 corridor, and as such will have a small advantage on radiational cooling nights.  Nights that are more advective favor the airport since it is further south.

Results:

After the data was compiled and the average, min, and max were calculated for each station as well as the differences between them, the summary sheet calculates the overall picture.

Looking over the station comparison above, the months where the cell color is red have large discrepancies and the yellow rows are for months that have fewer or smaller discrepancies.  The months where the data is complete without any highlighted discrepancies are in green.  Not that the difference shown for the green cells only to the right of the shaded columns is much more accurate and illustrates the tendency of the Ambient Weather 2902 to record higher highs (+0.7F) and lower lows (-1.0F) than the Davis Vantage Pro 2.  @Sabal_Louisiana mentioned this tendency above.

Of more concern for palm and tropical plant enthusiasts is how well each station performs with below average temperatures or freeze events.  While we did not have a record-setting event in the time frame recorded on both sensors, there was a record below freezing in January 2020.  Some of the other colder days were included for December, January, February, and March were included as well.  Note the similarities in the delta measurements under these more important conditions is nearly identical to the overall chart above:

Summary:

The Davis Vantage Pro 2 is an exceptionally accurate personal weather station (PWS).  That said, the Ambient Weather 2902 is certainly very close in regard to temperature and much less expensive.  The Ambient Weather tends to run +/- 0.7F when comparing temperatures with DVP2.  It is safe to say that if your Ambient Weather is in close proximity to a DVP2 and is within 1F of its records, it is likely correctly calibrated and reporting within reasonable parameters.  The speadsheet attached contains the weather records from WUnderground.com and the summary sheet with the information in the photos above.

20230701_AMB_vs_Davis_Stats.xlsx

[SeanK]

I had a LaCrosse for a couple years. Basic model, indoors, one additional Temp/RH sensor, internet connection.

I found it frustrating b/c I could not assign the local IP address, could not identify it in my router's list of connections, and it had no webpage. I buy HP printers and they show me their address, which has an actual webpage with information. 

Edited July 5 by SeanK

[JJPalmer]

This is great, Kinzy. Thanks for putting together.  Would love to know how some of the 'higher-end' Davis competitors fare, but it's hard to find the data.  I've been curious about the following stations:

• Ambient Weather WS-5000
  • Seems like it mostly updates the anemometer and rain gauge. Slightly annoying that they haven't created three separate sensors as you're supposed to have the anemometer mounted at 30' and thermometer at 6' or so, but alas. 
• Kestrel / Rainwise / Young, though these are very high-end and less common. 

[kinzyjr]

1 hour ago, JJPalmer said:

This is great, Kinzy. Thanks for putting together.  Would love to know how some of the 'higher-end' Davis competitors fare, but it's hard to find the data.  I've been curious about the following stations:

• Ambient Weather WS-5000
  • Seems like it mostly updates the anemometer and rain gauge. Slightly annoying that they haven't created three separate sensors as you're supposed to have the anemometer mounted at 30' and thermometer at 6' or so, but alas. 
• Kestrel / Rainwise / Young, though these are very high-end and less common. 

There is an Ambient WS5000 (KFLLAKEL443) in the same location as the DVP2 in the previous experiment report.  While there is another DVP2 close by (KFLLAKEL9), it has recorded extremely low numbers in the past and is either not reliable or the climate in that small area is just awful chilly.  An example is the 15.7F in January 2010.

The remaining DVP2 stations in the area that may be usable for comparison:

• KFLLAKEL41 (2.30 miles NW)
• KFLLAKEL38 (4.25 miles SW)
• KFLLAKEL25 (2.25 miles NW)
• KFLLAKEL22 (2.70 miles NNE)

There was a RainWise MK-III-LR station (KFLLAKEL283) in Lakeland at CODE Academy that has some data for December 2020 and January 2021.  The closest DVP2s would be south on the other side of Memorial Boulevard ( KFLLAKEL35 (data removed), KFLLAKEL32, KFLLAKEL172, KFLLAKEL259) or north of FL-33 (KFLLAKEL43, KFLLAKEL139).

I don't currently see any Kestrel or Young units in the area.

Certainly something to consider as long as there is some decent overlap in the dates each of the stations reported to Wunderground and/or AmbientWeather.net.

 

@SeanK

An interesting experience.  I haven't personally installed a LaCrosse, but wouldn't have expected them to be so difficult.  In the case of Ambient Weather 2902, I know that they require you to connect to the console's Wi-Fi (different network than SOHO routers: ex. 192.168.1.1 is the typical default for SOHO routers vs. 192.168.4.1 for AmbientWeather consoles).  In order to connect with a laptop, it typically involves powering them off and back on as they automatically stop advertising their Wi-Fi network beyond ~5 minutes for security purposes.  The Ambient Weather 2902 is the most common model purchased here. 

Did you replace the LaCrosse with another brand?

[SeanK]

17 hours ago, kinzyjr said:

There is an Ambient WS5000 (KFLLAKEL443) in the same location as the DVP2 in the previous experiment report.  While there is another DVP2 close by (KFLLAKEL9), it has recorded extremely low numbers in the past and is either not reliable or the climate in that small area is just awful chilly.  An example is the 15.7F in January 2010.

The remaining DVP2 stations in the area that may be usable for comparison:

• KFLLAKEL41 (2.30 miles NW)
• KFLLAKEL38 (4.25 miles SW)
• KFLLAKEL25 (2.25 miles NW)
• KFLLAKEL22 (2.70 miles NNE)

There was a RainWise MK-III-LR station (KFLLAKEL283) in Lakeland at CODE Academy that has some data for December 2020 and January 2021.  The closest DVP2s would be south on the other side of Memorial Boulevard ( KFLLAKEL35 (data removed), KFLLAKEL32, KFLLAKEL172, KFLLAKEL259) or north of FL-33 (KFLLAKEL43, KFLLAKEL139).

I don't currently see any Kestrel or Young units in the area.

Certainly something to consider as long as there is some decent overlap in the dates each of the stations reported to Wunderground and/or AmbientWeather.net.

 

@SeanK

An interesting experience.  I haven't personally installed a LaCrosse, but wouldn't have expected them to be so difficult.  In the case of Ambient Weather 2902, I know that they require you to connect to the console's Wi-Fi (different network than SOHO routers: ex. 192.168.1.1 is the typical default for SOHO routers vs. 192.168.4.1 for AmbientWeather consoles).  In order to connect with a laptop, it typically involves powering them off and back on as they automatically stop advertising their Wi-Fi network beyond ~5 minutes for security purposes.  The Ambient Weather 2902 is the most common model purchased here. 

Did you replace the LaCrosse with another brand?

It connected OK. When it worked it worked fine but I had no control over the connection. I never replaced it. 

With the HP printers, there is a USB port that lets you set the ethernet or wifi address.

[Barry]

I just installed a Weather Flow Tempest unit on my roof this weekend. It has good reviews. I like that it has no moving parts, we will see how it performs. 

[flplantguy]

21 hours ago, Barry said:

I just installed a Weather Flow Tempest unit on my roof this weekend. It has good reviews. I like that it has no moving parts, we will see how it performs. 

I'm very happy with mine so far. It was exactly accurate on the pressure reading when idalia passed and is accurate everywhere else as far as I can tell.  Rainfall is debatable since it's the first haptic rain sensor and I've seen mine adjust later, so if exact rain amounts is important all you need is a manual rain gauge and the rest is good. The app is amazing also and it shares the data with NOAA for forecasting models and such so your helping them too. Not sure yet on the individual forecast accuracy as that takes time but most people say they like it.

[Barry]

1 hour ago, flplantguy said:

I'm very happy with mine so far. It was exactly accurate on the pressure reading when idalia passed and is accurate everywhere else as far as I can tell.  Rainfall is debatable since it's the first haptic rain sensor and I've seen mine adjust later, so if exact rain amounts is important all you need is a manual rain gauge and the rest is good. The app is amazing also and it shares the data with NOAA for forecasting models and such so your helping them too. Not sure yet on the individual forecast accuracy as that takes time but most people say they like it.

Glad to hear you are happy with yours. I guess my debate would be to take it down or leave up in the event of a hurricane. I have it mounted pretty good, but not sure it would survive a direct hit like Ian was.

[JJPalmer]

Did some more digging into one of the newest PWS on market - the Kestrelmet 6000.  The station comes with a 24-hour fan-aspirated radiation shield, a propeller-style anemometer, and commercial grade equipment all for ~$1k with optional additional equipment.  From the reviews I've seen, this station is as-good-as or better-than the Davis Vantage Pro systems and is much easier to setup and view data (no additional equipment purchase necessary; connects directly to Ambient Weather).  Josh Morgerman - a professional hurricane chaser - just installed one on his new property in Mississippi in a ringing endorsement. 

 

If anyone is looking to purchase their first or upgrade their existing PWS monitor and were looking at Davis systems, I'd highly encourage you to check out the Kestrelmet 6000.   It's at the top of my list for when I eventually upgrade.  

[Xerarch]

Nice to see these comparisons, I have the Ambient Weather WS-2000 and have been pretty happy with it.  Once upon a time I had an Oregon Scientific and it was wildly inaccurate during the day, radiation shielding was not sufficient at all and it was not usable for daytime temperatures.  I was pleasantly surprised that my Ambeint Weather station gives me believable readings during the day even without being fan aspirated. By comparing to official readings and personal stations in the area I'm satisfied that it is accurate enough, at least considering my budget.

Of interest though, is that there is a Naval Air Station close to my house, in a similar site as my house, proximity to water etc.  It consistently records temperatures that are ~2-3 degrees warmer overnight than my station, also warmer than anyone's personal station in the area.  I don't get it, even during great equalizing events like the December 2022 freeze, that wind was so powerful that it overwhelmed any benefit from water or other microclimate, everyone's PWS read very close to others in the area, but somehow that Naval Station still read like 3 degrees warmer than mine.  I don't know if they have that station sitting inside a concrete box or what, but I have a hard time believing the reading when no others can seem to match it.

[kinzyjr]

3 hours ago, Xerarch said:

Nice to see these comparisons, I have the Ambient Weather WS-2000 and have been pretty happy with it.  Once upon a time I had an Oregon Scientific and it was wildly inaccurate during the day, radiation shielding was not sufficient at all and it was not usable for daytime temperatures.  I was pleasantly surprised that my Ambeint Weather station gives me believable readings during the day even without being fan aspirated. By comparing to official readings and personal stations in the area I'm satisfied that it is accurate enough, at least considering my budget.

Of interest though, is that there is a Naval Air Station close to my house, in a similar site as my house, proximity to water etc.  It consistently records temperatures that are ~2-3 degrees warmer overnight than my station, also warmer than anyone's personal station in the area.  I don't get it, even during great equalizing events like the December 2022 freeze, that wind was so powerful that it overwhelmed any benefit from water or other microclimate, everyone's PWS read very close to others in the area, but somehow that Naval Station still read like 3 degrees warmer than mine.  I don't know if they have that station sitting inside a concrete box or what, but I have a hard time believing the reading when no others can seem to match it.

There are issues like this with some of the other NOAA stations as well.  Just in the region, these two often have numbers that are well outside of what anything close by is recording:

KPCM: https://www.palmtalk.org/forum/topic/75122-florida-winter-2022-2023/?do=findComment&comment=1086979

KORL: https://www.palmtalk.org/forum/topic/75122-florida-winter-2022-2023/?do=findComment&comment=1088505

These are certainly not the only stations where this happens frequently.  About 15-20 years ago, there was a flaw in the ASOS stations that allowed heat to build up near the sensor, but that flaw should have been corrected in the stations currently in operation.  Aside from that, the records at KMCO are recorded with an ASOS unit and they are congruent with what is recorded locally in that area.  Too bad red liquid thermometers don't have automated reporting... at least not that I'm aware.  They tend to be a bit lower than electronic instruments of most brands.
 

3 hours ago, JJPalmer said:

Did some more digging into one of the newest PWS on market - the Kestrelmet 6000.  The station comes with a 24-hour fan-aspirated radiation shield, a propeller-style anemometer, and commercial grade equipment all for ~$1k with optional additional equipment.  From the reviews I've seen, this station is as-good-as or better-than the Davis Vantage Pro systems and is much easier to setup and view data (no additional equipment purchase necessary; connects directly to Ambient Weather).  Josh Morgerman - a professional hurricane chaser - just installed one on his new property in Mississippi in a ringing endorsement. 

 

If anyone is looking to purchase their first or upgrade their existing PWS monitor and were looking at Davis systems, I'd highly encourage you to check out the Kestrelmet 6000.   It's at the top of my list for when I eventually upgrade.  

Thank you for the recommendation.  I'm going to give this one a look.

[flplantguy]

I was just reading about the tempest stations and their firmware having algorithms that adjust for conditions and the sensor housing to determine the "real" temperature.  I will be setting another sensor with it now to measure the accuracy of it relative to that and the other sensors I know work well.  It may explain why all the sensors were at the same readings as the local stations, but the tempest (and the coldest sensor that's always lower) were 5 degrees cooler than the rest and all the local readings. It matches the areas slightly inland though, which were a degree or two colder-but also home stations in a cold microclimate.  I'm not sure which to trust so I will be observing it to see.  It picks up the sea breeze really well though, but still seems to run lower than the others nearby and with higher humidity levels in the afternoons.  It gets hit by sun all day after it clears the trees, and is open sky above all around at night, so it's very exposed relative to forest.