Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

Question regarding Lambert projection

William.Hatheway

Well-known member
I found a little disconnect between the online tutorial PowerPoints and the users guide online.

Does Lambert projection true lat values need to be fixed at 30 and 60 degrees or can they vary?
 
I found a little disconnect between the online tutorial PowerPoints and the users guide online.

Does Lambert projection true lat values need to be fixed at 30 and 60 degrees or can they vary?
I have been testing different configurations out and i don't see much of a difference but I could use some clarifying
 
I found a little disconnect between the online tutorial PowerPoints and the users guide online.

Does Lambert projection true lat values need to be fixed at 30 and 60 degrees or can they vary?
No, truelat for lambert projection can be changed based on your specific domain setting. They don't have to be 30N and 60N.
 
Hi,

I’m setting up a WRF simulation with the Lambert Conformal projection, and I’m trying to determine the correct values for truelat1 and truelat2 in the namelist.input file. Could anyone explain how these values are derived based on the domain's geographic properties (e.g., central latitude and longitude, and the area of interest)?

Thanks in advance for your help!
 
Lambert projection is appropriate for mid-latitude region, for example over 20-60 latitude north and south. There is no strict rule how to specifiy truelat1 and truelat2. One can determine their values based on the domain location and size. Please take a look at the document here, which describes how it works in WRF.
 
Lambert projection is appropriate for mid-latitude region, for example over 20-60 latitude north and south. There is no strict rule how to specifiy truelat1 and truelat2. One can determine their values based on the domain location and size. Please take a look at the document here, which describes how it works in WRF.
I am curious if, for the Lambert map projection, "ref_lat," "truelat1," and "truelat2" can be the same. It seems they shouldn't be, but each time I use WRFDomainWizard, the values for these parameters, as well as for "ref_lon" and "stand_lon," are the same, which doesn't seem logical.
 
"ref_lat"," "truelat1," and "truelat2" can be same. However, it is better to check map_factor to make sure it is around 1.0. Too large or too small value implies severe grid distortion, which is not expected.
 
Your truelat1 and truelat2 are set to the exact same value ($31.473$). While this works, it creates a "Tangent" projection that is most accurate only at that exact line.

Suggestion: To make the map accuracy more even across the whole state of Himachal, you could set truelat1 = 30.5 and truelat2 = 32.5. This spreads the "perfect" zones across the southern and northern parts of your domain.


Setting two different values for truelat1 and truelat2 (called the Secant case) is a trick to make your entire map more accurate, rather than just one single line.

Here is the simple breakdown of why this is better for Himachal Pradesh:

1. The "Touch" vs. "Cut" Method​

When you use a Lambert projection, you are basically placing a "cone" over the Earth.

  • Tangent (One True Lat): Your current setting (31.473 for both) means the cone just touches the Earth at that one line. The map is perfect at that line, but as you move North toward Lahaul or South toward Sirmaur, the distortion increases quickly.
  • Secant (Two True Lats): When you set them to 30.5 and 32.5, you are making the cone cut through the Earth. It enters at 30.5 N and comes back out at 32.5 N.

2. The "Sweet Spot" is Bigger​

By using two lines, you spread the accuracy across the whole state:

  • At 30.5°N: Error is 0%.
  • Between 30.5° and 32.5°: Error is extremely low (slightly compressed).
  • At 32.5°N: Error is 0%.
Because Himachal Pradesh is about 300 km tall (from south to north), using two lines ensures that the Map Scale Factor stays much closer to 1.000 for the entire domain. If you only use one line, the top and bottom of your map will have higher scale errors.

3. Impact on your 1km Nest​

Since you are going down to 1km resolution, you are trying to be very precise.

  • If your map scale factor is 1.002 instead of 1.000 because of distortion, your 1km grid cell is actually off by 2 meters every kilometer.
  • Over a 100km domain, that adds up to a 200-meter shift.
  • In the steep valleys of Kullu or Mandi, 200 meters is the difference between being at the bottom of a river bed or halfway up a mountain slope.

Summary:​

Setting truelat1 = 30.5 and truelat2 = 32.5 "brackets" the state of Himachal, ensuring that your high-resolution nests (3km and 1km) are sitting in the most mathematically stable part of the projection.
Does this make sense ?
@Ming Chen @William.Hatheway
 

Attachments

  • namelist.wps
    1.2 KB · Views: 0
Last edited:
Top