Working on tropical region, it is hard to find any day during the year that has no presence of a cloud. For instance, using the CM SAF product for year 2005 over the Lake Victoria, the least overpass of clouds, over a pixel of 4km horizontal resolution, was 4 on an hourly time scale. This means that each pixel is at least covered 17% of the time by a cloud for each day of the year.
Since my interest is to find "golden clear sky" conditions for the study of surface temperatures, I wonder how fast and how long does a cloud impact a surface temperature. Of course, this depends on multiple parameters, but what would commonly be accepted as a sufficient tampon time zone for minimizing the direct and retroactive impact of clouds on surface temperature?
Any literature or insights welcome.
Hi Oscar,
Crudely I would state that the prime variables to be considered are:
1) The speed of the cloud field;
2) How many octants it spans in the hemisphere en hence the frequency of direct and diffuse sunlight and their time spans of irradiation of the Earth's surface.
3) The thickness of the cloud layer(s) and its variation to calculate the magnitude of transmitted sunlight.
4) The thermal inertia of the soil on which solar irradiation direct or diffuse impinges.
Ideally, these variables will allow you for a reference point to calculate how fast and how long a cloud has a full impact on the SST of that point, hence when clear sky conditions occur over the point of reference.
Hi Oscar,
Crudely I would state that the prime variables to be considered are:
1) The speed of the cloud field;
2) How many octants it spans in the hemisphere en hence the frequency of direct and diffuse sunlight and their time spans of irradiation of the Earth's surface.
3) The thickness of the cloud layer(s) and its variation to calculate the magnitude of transmitted sunlight.
4) The thermal inertia of the soil on which solar irradiation direct or diffuse impinges.
Ideally, these variables will allow you for a reference point to calculate how fast and how long a cloud has a full impact on the SST of that point, hence when clear sky conditions occur over the point of reference.
Dear Oscar,
Further to the very pertinent answer provided by Frank, I'd like to add that while Equatorial regions are generally cloud-covered throughout the year, due to the raising branch of the Hadley Cell,
- in most regions, including within the tropics, there are strong seasonal variations, with more clouds during the rainy season and fewer during the dry season, linked to the latitudinal oscillation of the ITCZ, and
- especially in tropical regions, cloudiness exhibits a noticeable variation during the day, whereby clouds build up during the morning and throughout the day, and result in heavy precipitations in late afternoon and early evening.
Hence, if you are looking for the best opportunities to 'find "golden clear sky" conditions' in an area such as Lake Victoria, consider focusing on the early morning of days well within the 2 dry seasons.
Of course, another (complementary) approach might be to consider using data from other sources, and in particular satellites with a finer spatial resolution (e.g., 300 m or less), which will allow you to "peek" through and observe the land in the holes of a broken cloud field, if that is of use for your purposes.
Best regards, Michel.
Dear Oscar
several years ago, we made a simulator (OSIRIS) for simulating landscapes in thermal IR at very high resolution. We used this simulator to study the effect of the shades (that could be created by clouds) and found that the effect may last several hours after the disappearance of the shading object.
See
Article OSIrIS: a physically based simulation tool to improve traini...
(being uploaded)orArticle Influence of the three-dimensional effects on the simulation...
Best regards
Lucien
Dear all,
Thank you for your valuable answers. I will comment upon them once I have defined a proper methodology to find the "golden days" !
Kind regards,
Oscar
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing