Atmospheric River (AR) Detection using IPART

Atmospheric River (AR) Detection using IPART

1. What’s this about?

An atmospheric river (AR) in the field of meteorology/climatology refers to enhanced water vapor content (transport) in the lower troposphere.

Appearing as narrow and elongated vapor filaments in satellite imagery or gridded data, they represent transient and intensified horizontal moisture fluxes (Gimeno, Luis and Nieto, Raquel and V{’a}zquez, Marta and Lavers, David A., 2014, Dettinger, Michael, 2011).

A typical atmospheric river can carry 7-15 times the water in the Mississippi River (Ralph, F. M. and Neiman, P. J. and Kiladis, G. N. and Weickmann, K. and Reynolds, D. W., 2011), and at any time in winter, there are four to five such systems in the Northern Hemisphere alone (Zhu, Yong and Newell, Reginald E., 1998), accounting for 80-90% of the total north-south integrated vapor transport (Guan, Bin and Waliser, Duane E., 2015, Zhu, Yong and Newell, Reginald E., 1998).

In many existing applications, a magnitude-thresholding approach is used to detect such features from gridded data. For instance, (Ralph, F. Martin and Neiman, Paul J. and Wick, Gary A., 2004, Neiman, Paul J. and Ralph, F. Martin and Wick, Gary A. and Lundquist, Jessica D. and Dettinger, Michael D., 2008, Hagos, Samson and Leung, L. Ruby and Yang, Qing and Zhao, Chun and Lu, Jian, 2015) and (Dettinger, Michael, 2011) identified ARs by first locating regions where the Integrated Water Vapor (IWV) is greater than 20 mm. A 250 kg/(m (cdot) s) threshold on the Integrated Vapor Transport (IVT) was used by (Rutz, Jonathan J. and Steenburgh, W. James and Ralph, F. Martin, 2014, Rutz, Jonathan J. and Steenburgh, W. James and Ralph, Fred Martin, 2015).

However, an implicit assumption with this magnitude thresholding approach is that the atmospheric moisture level stays unchanged throughout the analysis period. Such an assumption may not be fully justifiable under a warming climate as the atmospheric moisture level is expected to increase.

This page shows an alternative detection method: Image-Process based Atmospheric River Tracking (IPART).

Instead of thresholding the IVT field “vertically” along the IVT magnitude dimension, we instead threshold on the “horizontal” scales of the feature, i.e. how wide such filaments are in space (< 1000 km) and how transient they are in time (< a couple of days).

Doing this allows the detection method to be less sensitive to the IVT intensity variations, and this insensitivity might be a crucial prerequisite for applications like climate projection inter-comparisons, where different models may have different levels of background IVT levels, or are modeling different future scenarios in which the moisture levels are supposed to differ notably.

The following part will cover:

  1. How the detection process work?
  2. IPART: A Python package implementing the algorithm
  3. Key parameters
  4. Input data
  5. Latest detections over the North Pacific and North Atlantic oceans
  6. Archived detections of the recent history

2. How the detection process works?

3. IPART: A Python package implementing the algorithm

IPART is also a Python package that handles the detection of ARs from gridded data. More specifically, it:

  1. Segments/masks out the regions of ARs from snapshots of IVT data.
  2. Finds an axis for each detected AR, as a simplified linear representation of the AR, and its major vapor transport direction as well.
  3. Tracks multiple instances of ARs across consecutive time points to for a track.

The code is maintained at Github: https://github.com/ihesp/IPART, with Jupyter notebooks examples within.

A more in-depth documentation of the package: https://ipart.readthedocs.io/en/latest/

The package is published at JOSS:

status.svg

4. Key parameters

Here are a list of the most important parameters used in generating the outputs down below:

  • kernel= [kt, ky, kx] = [0, 8, 8]. kt is the half length of the THR kernel in time dimension, kt=0 means it works on independent time slices. ky=kx=8 is the half length of the THR kernel in x- and y- dimensions. The total length is 2 * 8 + 1 = 17 grids, corresponding to 17 * 50 = 850 km, using data that have been regridded to (0.5 times 0.5 ^{circ}) resolution.
  • 'thres_low'= 45: (kg/(ms)), a preliminary IVT filtering threshold to remove low intensity noises.
  • 'min_area'= 50*1e4: (km^2), drop AR candidates smaller than this area.
  • 'max_area'= 1800*1e4: (km^2), drop AR candidates larger than this area.
  • 'min_LW'= 2, minimum required length/width ratio.
  • 'min_lat'= 20: degree, exclude systems whose centroids are lower than this latitude.
  • 'max_lat'= 80: degree, exclude systems whose centroids are higher than this latitude.
  • 'min_length'= 2000: (km), minimum required length of the AR.

5. Input data

The data used in generating such detections are:

  • Vertical integral of northward water vapour flux: u-flux.
  • Vertical integral of eastward water vapour flux: v-flux.

taken from the ERA5 reanalysis.

They are in (kg/(ms)) units, and are regridded to a horizontal resolution of (0.5 times 0.5 ^{circ}).

6. Latest detections over the North Pacific and North Atlantic oceans

6.1. North Pacific

north_pacific_AR_2023-01-14 18:00:00.png

6.2. North Atlantic

north_atlantic_AR_2023-01-14 18:00:00.png

7. Archived detections of the recent history

7.1. North Pacific

Click to see list 👇

7.2. North Atlantic

Click to see list 👇

8. References

Dettinger, Michael (2011). {Climate Change, Atmospheric Rivers, and Floods in California – A Multimodel Analysis of Storm Frequency and Magnitude Changes}, Journal of the American Water Resources Association.

Gimeno, Luis and Nieto, Raquel and V{’a}zquez, Marta and Lavers, David A. (2014). {Atmospheric rivers: A mini-review}, Frontiers in Earth Science.

Guan, Bin and Waliser, Duane E. (2015). {Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies}, Journal of Geophysical Research: Atmospheres.

Hagos, Samson and Leung, L. Ruby and Yang, Qing and Zhao, Chun and Lu, Jian (2015). {Resolution and Dynamical Core Dependence of Atmospheric River Frequency in Global Model Simulations}, Journal of Climate.

Neiman, Paul J. and Ralph, F. Martin and Wick, Gary A. and Lundquist, Jessica D. and Dettinger, Michael D. (2008). {Meteorological Characteristics and Overland Precipitation Impacts of Atmospheric Rivers Affecting the West Coast of North America Based on Eight Years of SSM/I Satellite Observations}, Journal of Hydrometeorology.

Ralph, F. M. and Neiman, P. J. and Kiladis, G. N. and Weickmann, K. and Reynolds, D. W. (2011). {A Multiscale Observational Case Study of a Pacific Atmospheric River Exhibiting Tropical–Extratropical Connections and a Mesoscale Frontal Wave}, Monthly Weather Review.

Ralph, F. Martin and Neiman, Paul J. and Wick, Gary A. (2004). {Satellite and CALJET Aircraft Observations of Atmospheric Rivers over the Eastern North Pacific Ocean during the Winter of 1997/98}, Monthly Weather Review.

Rutz, Jonathan J. and Steenburgh, W. James and Ralph, F. Martin (2014). {Climatological Characteristics of Atmospheric Rivers and Their Inland Penetration over the Western United States}, Monthly Weather Review.

Rutz, Jonathan J. and Steenburgh, W. James and Ralph, Fred Martin (2015). {The Inland Penetration of Atmospheric Rivers over Western North America: A Lagrangian Analysis}, Monthly Weather Review.

Zhu, Yong and Newell, Reginald E. (1998). {A Proposed Algorithm for Moisture Fluxes from Atmospheric Rivers}, Monthly Weather Review.

Author: guangzhi

Created: 2022-10-02 Sun 22:08

Validate