All-sky DNI exceeds clear-sky DNI in polar winter

Please enter here to ask a question about any NASA Science related topics!
Post Reply
james_newlands
Posts: 2
Joined: Wed Aug 25, 2021 1:03 am America/New_York

All-sky DNI exceeds clear-sky DNI in polar winter

by james_newlands » Wed Aug 25, 2021 1:17 am America/New_York

Hello all,

I'm using the POWER project data to estimate solar power budgets for glaciological instruments deployed in east Antarctica.

Looking at the hourly clear-sky and all-sky DNI data (CLRSKY_SFC_SW_DNI and ALLSKY_SFC_SW_DNI) I noticed that the all-sky value significantly exceeds the clear-sky value for one hour of several days either side of the mid-winter period with zero sunshine hours. Hopefully the attached graphs illustrate this.

This is at lat/lon -67.23, 114.47 (on the Totten glacier).

Is there any explanation for this behaviour?

Regards,

James
Attachments
clearsky_vs_allsky_zoom.png
clearsky_vs_allsky_zoom.png (88.29 KiB) Not viewed yet
clearsky_vs_allsky_full.jpg
clearsky_vs_allsky_full.jpg (352.14 KiB) Not viewed yet

Tags:

ASDC - David W.
User Services
User Services
Posts: 7
Joined: Tue Jun 08, 2021 11:29 pm America/New_York

Re: All-sky DNI exceeds clear-sky DNI in polar winter

by ASDC - David W. » Wed Aug 25, 2021 4:03 pm America/New_York

Thank you for your question. A Subject Matter Expert has been notified and will answer your question shortly. Please stand by!

ASDC - pstackhouse
Subject Matter Expert
Subject Matter Expert
Posts: 14
Joined: Thu Apr 02, 2020 4:36 pm America/New_York
Answers: 2
Location: NASA LaRC
Contact:

Re: All-sky DNI exceeds clear-sky DNI in polar winter

by ASDC - pstackhouse » Thu Aug 26, 2021 4:37 pm America/New_York

Dear James,

Given your latitude and the snow/ice surface, I think you found an inconsistency in the data that we will have to study more closely. The source data product for the solar irradiance on the horizontal surface (or GHI) is from the CERES SYN1Deg product and that GHI product is used to estimate the DNI. However, it appears that even the GHI all-sky exceeds the clear-sky for your location. This actually can physically happen due to multiple reflection between clouds and ice surface. The clear-sky flux wouldn't take that effect into account. Since our DNI is parameterized using the estimated direct energy on a horizontal surface and the solar weighted solar zenith angle during that period, I'm thinking that in your extreme circumstance the clear-sky horizontal irradiance is underestimated relative to the all-sky. Do you have any way to determine whether this happens under all conditions or under certain conditions like partially cloudy or cloud-free? We do try to provide estimated data quality information in our "methodology" and I will point out that we have less quality at high latitudes. Nevertheless, you found a condition that we'll have to evaluate more carefully to see if there's a way to provide you some additional guidance and/or impose some tests to prevent the inconsistency you are finding. Thank you for bringing this to our attention.

Sincerely,

Paul Stackhouse and The POWER Team
Dr. Paul Stackhouse
NASA Langley Research Center
Hampton, VA, USA

james_newlands
Posts: 2
Joined: Wed Aug 25, 2021 1:03 am America/New_York

Re: All-sky DNI exceeds clear-sky DNI in polar winter

by james_newlands » Mon Aug 30, 2021 3:03 am America/New_York

Hi Paul,

Thank you for your response.

Unfortunately I haven't had much luck finding Antarctic irradiance data (e.g. pyranometer measured GHI) to validate the CERES SYN1Deg product.

I'm mostly interested in worst case irradiances to size the battery bank for over winter operation, so I should be able to proceed by taking the minimum of allsky and clearsky irradiances from POWER / CERES SYN1Deg.

Regards,

James

ASDC - pstackhouse
Subject Matter Expert
Subject Matter Expert
Posts: 14
Joined: Thu Apr 02, 2020 4:36 pm America/New_York
Answers: 2
Location: NASA LaRC
Contact:

Re: All-sky DNI exceeds clear-sky DNI in polar winter

by ASDC - pstackhouse » Mon Aug 30, 2021 3:27 pm America/New_York

Dear James,

That sounds like a good strategy for safety, but I hope the underestimation won't cause too much problems in terms of cost.

Have you tried the BSRN archive for surface measurements. Two sites come to mind as relevant: Syowa (SYO; -69.005 lat, 39.589 lon), Georg Van Neumeyer (GVN; -70.65 lat, -8.25 long) and Concordia Station or DOME-C (DOM; -75.100 lat, 123.383 lon). Data from both sites in the BSRN archive at: https://bsrn.awi.de/. I hope these might help if you have found these.

We'll keep you mind as we investigate the issue over the next few weeks.

Best Regards,

Paul and the POWER Team
Dr. Paul Stackhouse
NASA Langley Research Center
Hampton, VA, USA

Post Reply