DataKind-DC / audubon-cbc

For the bird counters
MIT License
11 stars 18 forks source link

What do the Meteorologists say #47

Open rectheworld opened 4 years ago

rectheworld commented 4 years ago

Preform a Lit Review and Find data (with sources) to help answer the following question.

Are there any model's or analysis we should consider for our project hat you have seen in other articles or blogs?

jellena commented 4 years ago

Been trying dig into these a bit and the answer for all of them really seems to be that it really just depends.

What distances do meteorologists consider too far to be reliable?

There is no universal distance measure since factors such as ecosystem play such a large role in temperature variability. Temperature may be constant over long distance in environments such as a tundra or desert but less so for a coastal rain forest.

What elevations do Meteorologists consider to different to be reliable?

Similar issues arise as when comparing distance. However did see one blog that mentions for every 100 meters of altitude the temperature decreases by about 1 degree Celsius. But of course it also mentions situations of Temperature Inversion where that rule goes out the window.

https://sciencing.com/info-8686864-latitude-altitude-affect-temperature.html

What factors do meteorologists look at when comparing temperature, rain, ect?

In a perfect world they look at everything since it all seems to combine together. When researching the usability of citizen science I came across this paper that describes applying offsets to records after calibrating instruments in a "climate chamber" so the answer to the question is a meteorologist will look at as much as possible.

https://www.sciencedirect.com/science/article/abs/pii/S2212095517300068


Have we calculated the average distance and height differences between circles and the NOAA stations? Just trying to figure out the scope of the problem here.

rectheworld commented 4 years ago

Resources For Time Series Resources Found by Previous Volunteers

CALCULATING REGIONAL CLIMATIC TIME SERIES FOR TEMPERATURE AND PRECIPITATION: METHODS AND ILLUSTRATIONS

Link: https://rmets.onlinelibrary.wiley.com/doi/pdf/10.1002/(SICI)1097-0088(199604)16:4%3C361::AID-JOC53%3E3.0.CO;2-F

Not a paper but a resource William shared with us:

https://facebook.github.io/prophet/

From the site: "Prophet is a procedure for forecasting time series data based on an additive model where non-linear trends are fit with yearly, weekly, and daily seasonality, plus holiday effects. It works best with time series that have strong seasonal effects and several seasons of historical data. Prophet is robust to missing data and shifts in the trend, and typically handles outliers well."

Resource From Arch GIS - Python Notebook

https://notebooks.esri.com/user/z5uC66A3Uu5pEfGKkbRgI52bx/notebooks/samples/04_gis_analysts_data_scientists/temperature_forecast_using_time_series_data.ipynb

Summary: Uses Time Series analysis and LSTM (Long short-term memory), uses past temperature station data of weather stations in England to predict the next 30 days.