No Sun Days

Use this Forum to find information on, or ask a question about, NASA Earth Science data.
Post Reply
josh33
Posts: 2
Joined: Thu Jul 21, 2022 11:51 am America/New_York

No Sun Days

by josh33 » Thu Jul 21, 2022 12:31 pm America/New_York

I'm spec'ing off grid solar installations that need to remain online 24/7. I've been using the 'No Sun Days' from the POWER api to determine the number of consecutive days my installations will need to rely on batteries so I can determine bank sizing.

However, the parameter and data have changed on me twice now and I'm not sure which data is actually correct.

Examples below using 32.7767, -96.7970 (Dallas).

In 2019 the parameter was called "EQVLNT_NO_SUN_BLACKDAYS_MONTH". Worst month October (8.54 Days)

In late 2021 or early 2022 my existing api call no longer worked and the new parameter became "NO_SUN_BLACKDAYS_MAX". This time it only returned integers. Worst month March (5 Days)

Today my api call no longer works. The new parameter seems to be "MAX_EQUIV_NO_SUN_DAYS". Worst month as October (10.67 Days)

I don't mind maintaining the api call and updating the parameter every year but the difference between 5 days and 10.67 days of batteries is quite a lot.

Thanks for any insight you can offer.

Tags:

ASDC - cheyenne.e.land
User Services
User Services
Posts: 75
Joined: Mon Mar 22, 2021 3:55 pm America/New_York
Answers: 1
Been thanked: 3 times

Re: No Sun Days

by ASDC - cheyenne.e.land » Fri Jul 29, 2022 11:27 am America/New_York

Hello,

We apologize for the delay, a Subject Matter Expert has been notified and will be in contact with you shortly. Thanks.

ASDC User Services

ASDC - cheyenne.e.land
User Services
User Services
Posts: 75
Joined: Mon Mar 22, 2021 3:55 pm America/New_York
Answers: 1
Been thanked: 3 times

Re: No Sun Days

by ASDC - cheyenne.e.land » Wed Aug 10, 2022 4:07 pm America/New_York

Hello we apologize for the delay, but thank you for using POWER,

The Services Version 2 of POWER provides Data Version 901 POWER data (Services Version 1 provided Data Version 801). The Data Version 901 improves the 1) underling source data models and 2) enhances traceability to source data models. Specifically, the underling source data models for radiation-based parameters has changed from SRB v3 to SRB v4 for 1984 through 2000. From 2001, we now provide CERES SYN1deg until three months before present time. Like in previous versions, POWER continues to provide support for Near Real Time (NRT) availably with FLASHFlux. For the enhanced traceability, Data Version 901 POWER data includes improved parameter abbreviations and definitions, in order to be more clear about what parameter is being used. The most prominent change is that POWER provides data at the underlying source data model's grid resolution.

So, changes in the values are to be expected:
  • For meteorology-based parameters, we are no-longer using bilinear interpolation to put the source data on a common 0.5 x 0.5 degree grid as it was in Data Version 801 (to enhance traceability). Not using interpolation causes some slight variation in the values compared to the previous version, but the values are exactly the same as the source dataset for a particular point location. The differences in values between Data Version 801 and 901 could be more pronounced for a specific point location if:
    • the location is close to a land/sea boundary.
    • the location is close to the grid box boundary (since the data could potentially have pulled from a adjacent grid box when it was not using original source data grid).
  • For solar-based parameters, we are no-longer replicating the source data on a common 0.5 x 0.5 degree grid; we changed the underling source data models to newer versions. For more information, please review the data sources here.
  • The elevation, or altitude, is also not interpolated so the same reasons for meteorology-based parameters apply.
For this parameter in specific, Equivalent no sun days is calculated using the formula:
Image
      
Converting Minimum Available Insolation Over a Consecutive n-day period to Minimum Available Insolation Over a Consecutive n-day period (%) uses the following formula:
Image

The difference you are experiencing is due to the revised parameter definitions and the formula that we have used. The parameter “EQUIV_NO_SUN_BLACKDAYS_MONTH” (2019), “NO_SUN_BLACKDAYS_MAX” (2021), and “MAX_EQUIV_NO_SUN_DAYS” (now) refer to the same parameter. The most recent values are the correct ones.

Regards,
ASDC User Services

josh33
Posts: 2
Joined: Thu Jul 21, 2022 11:51 am America/New_York

Re: No Sun Days

by josh33 » Fri Aug 12, 2022 5:25 pm America/New_York

ASDC - cheyenne.e.land wrote:
> Hello we apologize for the delay, but thank you for using POWER,
>
> The Services Version 2 of POWER provides Data Version 901 POWER data
> (Services Version 1 provided Data Version 801). The Data Version 901
> improves the 1) underling source data models and 2) enhances traceability
> to source data models. Specifically, the underling source data models for
> radiation-based parameters has changed from SRB v3 to SRB v4 for 1984
> through 2000. From 2001, we now provide CERES SYN1deg until three months
> before present time. Like in previous versions, POWER continues to provide
> support for Near Real Time (NRT) availably with FLASHFlux. For the enhanced
> traceability, Data Version 901 POWER data includes improved parameter
> abbreviations and definitions, in order to be more clear about what
> parameter is being used. The most prominent change is that POWER provides
> data at the underlying source data model's grid resolution.
>
> So, changes in the values are to be expected:
>
> [list]For meteorology-based parameters, we are no-longer using bilinear
> interpolation to put the source data on a common 0.5 x 0.5 degree grid as
> it was in Data Version 801 (to enhance traceability). Not using
> interpolation causes some slight variation in the values compared to the
> previous version, but the values are exactly the same as the source dataset
> for a particular point location. The differences in values between Data
> Version 801 and 901 could be more pronounced for a specific point location
> if:
>
> [list=2]the location is close to a land/sea boundary.[/list]
> [list]the location is close to the grid box boundary (since the data could
> potentially have pulled from a adjacent grid box when it was not using
> original source data grid). [/list][/list]
>
> [list]For solar-based parameters, we are no-longer replicating the source
> data on a common 0.5 x 0.5 degree grid; we changed the underling source
> data models to newer versions. For more information, please review the data
> sources
> [url=https://power.larc.nasa.gov/docs/methodology/data/sources/]here[/url].[/list]
> [list]The elevation, or altitude, is also not interpolated so the same
> reasons for meteorology-based parameters apply.
> [/list]
>
> For this parameter in specific, Equivalent no sun days is calculated using
> the formula:
>
> [img]https://i.postimg.cc/jCYrxx5R/Screen-Shot-2022-08-10-at-3-51-55-PM.png[/img]
>       
> Converting Minimum Available Insolation Over a Consecutive n-day period to
> Minimum Available Insolation Over a Consecutive n-day period (%) uses the
> following formula:
>
> [img]https://i.postimg.cc/sxYhmJMy/Screen-Shot-2022-08-10-at-4-06-36-PM.png[/img]
>
> The difference you are experiencing is due to the revised parameter
> definitions and the formula that we have used. The parameter
> “EQUIV_NO_SUN_BLACKDAYS_MONTH” (2019), “NO_SUN_BLACKDAYS_MAX” (2021), and
> “MAX_EQUIV_NO_SUN_DAYS” (now) refer to the same parameter. The most recent
> values are the correct ones.
>
> Regards,
> ASDC User Services

I really appreciate the detailed follow up. Was this always the formula used and if not is there any way of getting the older parameters back?

For reference I had the older definitions commented in my code -
EQUIV_NO_SUN_BLACKDAYS_MONTH” (2019) - Equivalent number of NO-SUN or BLACK days is the number of days that a backup storage system must provide energy to account for the expected days without solar insolation.

NO_SUN_BLACKDAYS_MAX” (2021) - The maximum number of consecutive days that a backup storage system must provide energy to account for the expected days without solar insolation

MAX_EQUIV_NO_SUN_DAYS” (now) - Maximum equivalent no sun days is the maximum or the worse no sun days among all consecutive n-day equivalent no-sun days in a particular month

Based on the older definitions, I was led to believe you were tracking consecutive no-sun days in a given month (which is useful for spec'ing battery bank sizes) whereas now it's a total number of no-sun days over a consecutive period (which would tell the total number of days in a month that the battery bank would run in a deficit but not necessarily consecutively?).

Again, thank you for your help
Last edited by josh33 on Fri Aug 12, 2022 5:27 pm America/New_York, edited 1 time in total.

ASDC - c_mikovitz
Subject Matter Expert
Subject Matter Expert
Posts: 2
Joined: Mon May 03, 2021 6:39 am America/New_York

Re: No Sun Days

by ASDC - c_mikovitz » Mon Aug 15, 2022 2:47 pm America/New_York

Hello,

We realized that the previous formula was incorrect and not giving the desired results, therefore the previous method will not be available. The parameter was never intended to be actual consecutive no-sun days, but equivalent no-sun days. This work is based off a presentation at the Solar World Conference in 2005 by Whitlock, et al., which we will attach here. Hopefully this will clear up your questions, but if you need further clarification, please don't hesitate to ask.
Whitlock_power_sys.pdf
(163.86 KiB) Downloaded 19 times

josh33
Posts: 2
Joined: Thu Jul 21, 2022 11:51 am America/New_York

Re: No Sun Days

by josh33 » Thu Aug 18, 2022 6:08 pm America/New_York

Thank you for sharing this paper. The methodology is much clearer now. As I understand it, there can be a consecutive day deficit below average, however, the total insolation over that period could still be sufficient to keep the system online, depending on the panel sizing. It seems that the Equivalent No-Sun Days won't directly help spec our battery bank sizes (like the paper suggests), and I'll have to consider the insolation data much more closely.

Thank you again!

Post Reply