3
$\begingroup$

I am wondering if there is long term (10 year, 7 year, 20 year) periodic patterns Similar to short term seasonal and periodic patterns in weather such as sea/land wind, monsoon, ENSO, ocean sea-saw, etc.
Specifically related to drought
I am trying to predict the rainfall using the weather data, I was wondering if different patterns that happen in weather changes through time are periodic? Is there a periodic pattern in the pattern of weather changes? Is there a general periodic pattern that may happen every 10 years or more? If so, how long is the window of this period? Or a similar question can be: the weather measurements of how far ago can affect today's weather measurements? Does the weather forget about the weather events of long time ago? If so, how long is that window?

If there is a journal paper or material that I can study please tell me. Thanks

Additional Explanation: I am a data scientist trying to predict rainfall using weather data. So I do not have earth science knowledge. So after I read @Fred's comment, I understood that what I have in mind is called 'Cycle' in earth sciences.
edited question: What is the length of the cycles for rainfall patterns? (If I want to predict rainfall, how long is the length for cycles that affect rainfall? Or how many years of data I should use?! )

Improve this question
$\endgroup$
10
  • 4
    $\begingroup$ have you done any steps to find information about this subject,it might help you in making more focused questions. $\endgroup$
    – trond hansen
    Commented Jul 3, 2020 at 22:18
  • $\begingroup$ @a_donda Based on the title only, one could tell a little something about solar cycles and Milankovitch cycles. $\endgroup$
    – Joscha Fregin
    Commented Jul 4, 2020 at 0:21
  • $\begingroup$ With climate, many thing come into play. As others have stated, Milankovitch Cycles, resulting from the Earth's axial, precession & eccentricity creates climate cycles that have cycles for 1000s of years. Additionally, amongst other cycles, there are also drought cycles, with shorter intervals. You might find Climate: The Key To Understanding Business Cycles interesting, if you can find a copy. ... $\endgroup$
    – Fred
    Commented Jul 4, 2020 at 1:33
  • 1
    $\begingroup$ I can't tell if you are asking about climate or weather and there are several different questions here. Can you narrow this down to something a little more specific? I think this is a great line of questioning and you can always ask more questions, but it's better to keep each one well-focused so that it can have a fairly short, clear answer if possible. Thanks! $\endgroup$
    – uhoh
    Commented Jul 4, 2020 at 3:18
  • $\begingroup$ @trondhansen Yes, I have done, but I still do not have a more focused question. Since Earth Science is not my area of expertise and I am doing a project in this area, I needed some information. But I will explain my project in the descriptions so that maybe it clarifies what information I need. $\endgroup$
    – fof
    Commented Jul 4, 2020 at 13:18

1 Answer 1

Reset to default
0
$\begingroup$

I'll address this part of the question first:

"Or a similar question can be: the weather measurements of how far ago can affect today's weather measurements? Does the weather forget about the weather events of long time ago? If so, how long is that window?"

The answer is "approximately two weeks". Even with nearly perfect knowledge of the atmospheric state (the distribution of pressure, moisture, aerosols, etc), the chaotic nature of fluids means that beyond two weeks, one would be better of predicting the weather from a climatological estimate (i.e, the average of the weather on that date for the last ~30 years). Research into this very question was foundational for the field of Chaos theory (Lorenz, 1963 is a famous example). This isn't just an artifact of the predictive models, but the natural system itself. The weather has effectively 'forgotten' what it was doing a few weeks before.

While the atmosphere has a short 'memory', other parts of the Earth have longer 'memories'. Notable examples are the Oceans and the Landsurface. A warm anomaly in the Ocean will stick around for a while, since ocean currents are much slower than air for mixing things around (and diffusing out the anomaly). The atmosphere is very influenced by what happens on the surface, so the warm/cool anomaly can affects the weather in various ways. This is what happens in the case of ENSO that you have mentioned. The land can also have an impact. The amount of saturation in the soil can have quite a big impact on weather, especially for things like thunderstorms and monsoon systems. Continental snow cover extent may also impact atmospheric predictability. Stratosphere/troposphere coupling mechanisms might also be a source of longer term predictability.

And this part: "I am wondering if there is long term (10 year, 7 year, 20 year) periodic patterns Similar to short term seasonal and periodic patterns in weather such as sea/land wind, monsoon, ENSO, ocean sea-saw, etc."

The most well known ~10 year climate oscillation is the Pacific Decadal Oscillation.

Improve this answer
$\endgroup$
3
  • $\begingroup$ Thank you so much for detailed explanation. Is there a way to capture these patterns such as ENSO, etc in data analytics? Will they be captured automatically if I try to find the seasonalities within the data? $\endgroup$
    – fof
    Commented Jul 6, 2020 at 20:07
  • $\begingroup$ Well, start with predicting the next El Nino, start, peak, end. $\endgroup$
    – user20217
    Commented Jul 7, 2020 at 14:26
  • 1
    $\begingroup$ @fof-- The answer is 'it depends'. There might not be any correlation with between El Nino and precipitation whatsoever, depending on what part of world you're conducting your analysis. You probably want to start by looking at ENSO indices and seeing if your data correlates. origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/… $\endgroup$
    – William Rudisill
    Commented Jul 7, 2020 at 17:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.