# (Short-term) Forecasting or (Concurrent) Estimation?

**URL:** <https://community.drivendata.org/t/short-term-forecasting-or-concurrent-estimation/7299>\
**Category:** NASA Airathon\
**Created:** [February 11, 2022, 10:05am UTC](https://community.drivendata.org/t/short-term-forecasting-or-concurrent-estimation/7299 "2022-02-11T10:05:51Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![sukantabasu](https://avatars.discourse-cdn.com/v4/letter/s/e9a140/32.png) [@sukantabasu](https://community.drivendata.org/u/sukantabasu)\
**Post date:** [February 11, 2022, 10:05am UTC](https://community.drivendata.org/t/short-term-forecasting-or-concurrent-estimation/7299/1 "2022-02-11T10:05:51Z")

</div>

In the problem description, it has been mentioned that:

“Remember that for each observation, you may only use input values that are available **before** this time.”

“Remember that you are not allowed to use future data, so the time\_end of an granule must be **before** the datetime of a given observation.”

In other words: satellite\_time\_ **end** \< observation\_datetime\_ **start**

Thus, the problem is a short-term forecasting problem.

However, the information provided by **czsc** in the discussion forum is different:

“Therefore, you can use satellite data with an endtime on or before 2019-02-01T07:59:00Z for a label with datetime 2019-01-31T08:00:00Z.”

“In short, you are allowed to use data up-through (i.e. including) the date of prediction. This means that yes, you can use satellite data from the same local date.”

This statement implies: satellite\_time\_ **end** \< observation\_datetime\_ **end**.

Thus, according to **czsc** , the problem is a concurrent estimation problem. It is in contradiction to the problem statement.

If the satellite data are used from the same date as observation, then we are not doing a short-term (say day-ahead or hours-ahead) forecasting.

Please clarify.

---

<div class="post-metadata">

**Author:** ![cszc](https://avatars.discourse-cdn.com/v4/letter/c/c68b51/32.png) [@cszc](https://community.drivendata.org/u/cszc)\
**Post date:** [February 16, 2022, 9:00pm UTC](https://community.drivendata.org/t/short-term-forecasting-or-concurrent-estimation/7299/2 "2022-02-16T21:00:31Z")

</div>

Thanks for this questions @sukantabasu . Sorry about the confusion. It is correct to think about this problem as a concurrent estimation or “nowcasting” problem, i.e. satellite\_time\_ **end** \< observation\_datetime\_ **end**. We will clarify the language on the website.

---

<div class="post-metadata">

**Author:** ![future-forecast](https://avatars.discourse-cdn.com/v4/letter/f/82dd89/32.png) [@future-forecast](https://community.drivendata.org/u/future-forecast)\
**Post date:** [March 20, 2022, 3:53am UTC](https://community.drivendata.org/t/short-term-forecasting-or-concurrent-estimation/7299/3 "2022-03-20T03:53:49Z")

</div>

Hi!

> We will clarify the language on the website.  
> Not written. So please tell us Clearly.

1. When and how many days should we forcast?

2.How many points every point should we forcast?  
Los Angeles (South Coast Air Basin)  
Delhi  
Taipei

How many points at each point?

Please clarify.  
Thank you!

---

<div class="post-metadata">

**Author:** ![cszc](https://avatars.discourse-cdn.com/v4/letter/c/c68b51/32.png) [@cszc](https://community.drivendata.org/u/cszc)\
**Post date:** [March 20, 2022, 6:38pm UTC](https://community.drivendata.org/t/short-term-forecasting-or-concurrent-estimation/7299/4 "2022-03-20T18:38:20Z")

</div>

Hi @future-forecast. For details on inference, please refer to the `submission format` section of the problem description ([pm2.5](https://www.drivendata.org/competitions/91/competition-air-quality-no2/page/435/#sub) and [no2](https://www.drivendata.org/competitions/91/competition-air-quality-no2/page/435/#sub)). Your submission should include a prediction for every row in `submission_format.csv`, where each row represents a 24 hours period for a given grid cell.

> [@future-forecast](#):
>
> When and how many days should we forcast?

You should forecast the average concentration for the 24 hours following the timestamp. From the problem description:

> The UTC datetime of the measurement in the format `YYYY-MM-DDTHH:mm:ssZ` . A `value` represents the average between 12:00am to 11:59pm local time. The `datetime` provided represents the start of that 24 hour period in UTC time.

> [@future-forecast](#):
>
> 2.How many points every point should we forcast?

You should predict for every `grid_id` in the `submission_format.csv`. You can find more information about the grid cells in the `grid metadata` file on the `data download` page.

Hope that clarifies some things!