Hi
I'm trying to find a solution for a problem related to satellote temperature data.
I work with intertidal environment, and I have temperature sensores deployed in these environment for years. Data shows the temperature observed by these in situ sensors are quite diferent from the satellite.
The problem is, I intend to do mechanistic species distribution models, but my mechanistic data was based on data colected by the in situ sensors, and these only exist on a few places, whereas the satellite data is everywhere.
Assuming, my in situ sensors are the 'correct ones', is there a way to calibrate the satellite data for the places where I have no sensors according to the diferences observed between satellite and in situ sensors in the places I do have sensors?
Thank you.
Cheers, Luís Pereira.
Hi, these discrepancies could be addressed by several posibilities
1. Variable: Satellite by itself does not measure the air temperature, but brightness temperature, and estimate the land surface temperature (LST). Are your sensors measuring LST or air temperature? Depending of land cover characteristics, the difference could be important between those two variables.
2. Spatial resolution: does the cover where place is measured a representative cover to compare with the spatial resolution with satellite data are you working?
3. Uncertainities of measurement: If effectively are you working with LST, the uncertainities of these product are addressed depending of emissivity estimation and atmospheric conditions for the sensor measurement versus the satellite estimation.
4. If not and you are working with air temperature as a model from satellite data, depending on model you should check the uncertainities of that product.
Maybe this document can help to answer better your question
https://lpvs.gsfc.nasa.gov/PDF/CEOS_LST_PROTOCOL_Feb2018_v1.1.0_light.pdf
Thank you so much Italo Moletto-Lobos
Regarding your questions:
1- Yes the sensors are embebed in rock at the intertidal shore.
2- Well...hand't decide the grid I'll be using, but the sensors discriminate sites distanced at least 100km from each other.
3- Yes, I intend to use relevant factors like emissivity and atmospheric conditions to approximate the remote sensing methods (satellite fo instance) to the in site measurements.
4- Not sure if that's what I'm looking for.
Ideally, I would get satellite (and weather station data maybe) and use it as input for a model that predict in situ data for the sites where I have no sensors.
Article Assessing high resolution thermal monitoring of complex inte...
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity