<style>
table {
font-size: 10px
}
</style>
## Introduction
Modelling atmospheric dispersion and deposition of volcanic ash is becoming increasingly valuable for understanding the potential impacts of volcanic eruptions on infrastructures, air quality and aviation
| Pilot 12 | High-resolution volcanic ash dispersal forecast |
|:--------:| ----------------------------------------------------------- |
| BSC^1^ | A. Folch, L. Mingari, A. Prata |
| INGV^2^ | A. Costa, G. Macedonio, F. Pardini, M. De' Michieli Vitturi |
| IMO^3^ | S. Barsotti, a PostDoc |
^1^ Barcelona Supercomputing Center (BSC)
^2^ Istituto Nazionale di Geofisica e Vulcanologia (INGV)
^3^ Icelandic Meteorological Office (IMO)
| | | |
| -------- | -------- | -------- |
| Logo_bsc | Logo_ingv| Logo_imo |
---
## Objectives for PD12
* This PD will increase resolution of present operational ash dispersal model setups by one order of magnitude
* The pilot will consider an European domain including Iceland, i.e., (~4,000x4,000 km<sup>2</sup> of spatial coverage) with a horizontal resolution of about 4 km and 100 vertical levels
* Development an ensemble-based data assimilation system combining the FALL3D ash dispersal model with the Parallel Data Assimilation Framework (PDAF)
* Produce a joint estimation of 4D ash concentration forecasts and, simultaneously, an optimization of the eruption source term (e.g., injection column height or vertical mass distribution)
* Consider an ensemble containing a minimum of 30 members
* Run the following data assimilation test cases:
- Eyjafjallajökull (2010) and Grímsvötn (2011) eruptions (Iceland)
- Etna eruptions: 23rd February 2013 and 23rd November 2013 (Italy)
---
## Target users
* Technology transfer is expected towards weather services and world VAACs (<ins>V</ins>olcanic <ins>A</ins>sh <ins>A</ins>dvisory <ins>C</ins>enters)
* Meteorologists on duty for the identification of areas in the atmosphere that could be unsafe for aviation operations as well as on the ground (e.g. at the airports)
* Forecasting ash clouds during an ongoing eruption or quantifying potential impacts from future eruptions are relevant issues to aviation stakeholders and to civil protection agencies and governmental bodies
* Aviation sector and related stakeholders
---
## Current status
* [x] The **FALL3D** code has been redesigned and rewritten from scratch in the framework of the EU Center of Excellence for Exascale in Solid Earth (ChEESE)
* [x] The new version of FALL3D is tailored to the extreme-scale computing requirements
* [x] Ensemble forecasting. Future release (version 8.1)
* [x] Data insertion: Satellite-retrieved data from recent volcanic eruptions can be used as initial condition
Work in progress:
* [ ] Implement an ensemble-based data assimilation system in the new version 8 of the FALL3D model for volcanic ash and gases (e.g., SO<sub>2</sub>)
* [x] Build an assimilation system by coupling the numerical model **FALL3D** with the the Parallel Data Assimilation Framework (**PDAF**) into a single program
* [ ] Assimilation of multiple observational data sources, e.g., satellite retrievals of volcanic ash mass loadings, lidar/ceilometer data
---
## Data assimilation system (FALL3D+PDAF)
**What is FALL3D?**
* FALL3D is a model for atmospheric passive transport and deposition of particles, aerosols, and radionuclides.
* Originally developed for volcanic particles, has a track record of 50+ publications on different model applications and code validation, as well as an ever-growing community of users worldwide, including academia, private, research, and several institutions tasked with operational forecast of volcanic ash clouds
* Source code: [FALL3D](https://gitlab.com/fall3d-distribution/v8.0)
* Documentation: [Wiki](https://gitlab.com/fall3d-distribution/v8.0/-/wikis/home)
**What is PDAF?**
* PDAF is an open-source software environment for ensemble data assimilation providing fully implemented and optimized data assimilation algorithms
* Include ensemble-based Kalman filters
<center>
<img src="https://i.imgur.com/pVK9s4w.png" height="120">
<img src="https://i.imgur.com/jtb51Bw.png" height="120">
</center>
---
## Ensemble forecasting (I)
* Multiple simulations are run to produce a range of possible system states
* Objective: represent the uncertainty in numerical models
* Each run with perturbed model parameters and variation of its initial conditions
---
## Ensemble forecasting (II)
**Basic ensemble products**
* Postage Stamps: all the ensemble members can be displayed together for visual comparison
* Ensemble mean
* Ensemble probabilities
* Ensemble spread: a measure of the difference between the members. Large spread indicates a low forecast accuracy
| Ensemble Mean | Probabilities |
|:------------------------------------:|:------------------------------------:|
|  |  |
---
## Data assimilation
* The generation of high-quality forecasts depends on the accuracy and reliability of the input data for models
* Uncertainties in key parameters such as column height injection, physical properties of particles, or meteorological fields, represent a major source of error in forecasting airborne volcanic ash
* **Data assimilation** is one of the most effective ways to reduce the error associated with the forecasts through the incorporation of available observations into numerical models
Sequential assimilation:
- Correct model state estimate when observations are available (analysis)
- Propagate estimate (forecast)
---
## New generation of geostationary satellites
| Satellite | Sensor | Coverage | Spatial res. | Temporal res. | Ash/SO<sub>2</sub> bands (mum) | Lifetime |
| ----------- | ------ | --------------- | ------------ | ------------- | ------------------------------- | --------- |
| Meteosat-11 | SEVIRI | Europe/Africa | 3 km | 15 min | 7.35, 8.7, 10.8, & 12 | 2015-2022 |
| FY-4A | AGRI | S. Asia/Oceania | 4 km | 15 min | 8.5, 10.7, & 12 | 2016-2021 |
| Himawari-8 | AHI | S. Asia/Oceania | 2 km | 10 min | 7.35, 8.6, 10.45, 11.2, & 12.35 | 2014-2029 |
| GOES-17 | ABI | W America | 2 km | 10 min | 7.4, 8.5, 10.3, 11.2, & 12.3 | 2018-2029 |
| GOES-16 | ABI | E America | 2 km | 10 min | 7.4, 8.5, 10.3, 11.2, & 12.3 | 2016-2027 |
| Meteosat-11 | FY-4A | Himawari-8 | GOES-17 | GOES-16 |
| -------- | -------- | -------- | --- | --- |
|  | Text |  | img | img |
---
## Implementation of a data assimilation system
The numerical model FALL3D coupled with the the Parallel Data Assimilation Framework (PDAF) into a single program:
<center>
<img src="https://i.imgur.com/yag9sM3.png" height="400">
</center>
---
## Data assimilation: Example (I)
* Filter: ETKF. The Ensemble Transform Kalman Filter (ETKF) provides very efficient ensemble transformation
---
## Data assimilation: Example (II)
**Ensemble forecasting:**
---
## Data assimilation: Example (III)
**Assimilation step:**
---
## Data assimilation: Example (I)
* Raikoke volcanic eruption (2019) - Low resolution simulation
* Ensemble Transform Kalman Filter (ETKF)
* Observation data: SO<sub>2</sub> mass loading
* Satellite: HIMAWARI 8
* Ensemble size: 48
|Ensemble forecasting|Observation|Analysis|
|:------------------:|:---------:|:------:|
|  |  |  |
---
## Data assimilation: Example (II)
* Puyehue-Cordón Caulle eruption (2011) - Low resolution simulation
* Ensemble Transform Kalman Filter (ETKF)
* Observation data: volcanic ash mass loading
* Satellite: SEVIRI
* Ensemble size: 48
|Ensemble forecasting|Observation|Analysis|
|:------------------:|:---------:|:------:|
| |  |  |
---
## Future work
* Implement and test other filters
- For instance, LETKF: a localised form of the ETKF. The analysis and the ensemble transformation are performed in a loop through disjoint local analysis domains
* Use different metrics to assess the performance of the assimilation system: RSME, SAL...
* Perform high-resolution simulations for the identified test cases:
- Two recent eruptions with a good observational dataset available in Iceland: the Eyjafjallajökull eruption in 2010 and the Grímsvötn eruption in 2011
- Two recent eruptions of Etna (Italy) with available field, radar, ceilometer, and satellite data: the 23rd February 2013 and the 23rd November 2013 eruptions.
---
## Thank you!
{"metaMigratedAt":"2023-06-15T04:39:58.450Z","metaMigratedFrom":"YAML","title":"High-resolution volcanic ash dispersal forecast","breaks":true,"description":"2nd ChEESE General Assembly, Paris","contributors":"[{\"id\":\"8d65bcb6-dc49-4a01-a201-024e9af0f499\",\"add\":19949,\"del\":10514}]"}