Thread by @Dr_Gilbz

1/ Did somebody say NEW PAPERS???

The last part of my PhD work is published today in JGR Atmospheres (early online access) as a 2-part series.

👉 part 1: doi.org/10.1029/2021JD…
👉 part 2: doi.org/10.1029/2021JD…

@BAS_News @polarpredictian

Get ready for… a mega thread:

2/ We showcase a high-res, 20-year regional #ClimateModel hindcast forced w/ re-analysis that addresses lack of long-term obs in the region.

We use the dataset to quantitatively rank - for the 1st time - the most important drivers of melting on Larsen C.

3/ It’s a powerful tool to explore surface conditions, energy balance, near-surface #meteorology, #cloud properties and upper air flow.

But don’t take my word for it, here’s proof…

4/ Crucially, the MetUM is able to reproduce patterns of surface #melting captured by satellites, like this composite from Bevan et al. (2018).

The melt maps (MetUM left, sat right) show:
1. N-S gradient related to temp and SW radiation
2. E-W gradient related to #foehn winds

5/ We look at case studies to explore causes of melting in detail. 1: collapse of Larsen B in 2002.

Melting assoc w/ unusually warm conditions + foehn drove hi melt, esp. close to mountains, triggering #IceShelf collapse

Fig: Mean melt amnt in 8 months b4 Larsen B’s collapse

6/ 2: intense foehn-driven melt of MAM'16 (c.f. Kuipers Munneke et al., 2018). MAM'16 was a v foehn-heavy season, with 30x more melt than seas average, driven lg'ly by sensible heat.

Fig: Annual mean foehn occurrence 1998-2017 (left) and mean foehn occurrence MAM'16 (right)

7/ Great! The hindcast works well to explore the causes of melt. So next we use it to rank the most important causes of melting on the Larsen C ice shelf. So, I give you: the Top of the Melt charts.

8/ In first place: solar radiation (not a huge surprise). Sunny conditions cause widespread and considerable melting, especially in summer, when 90% of melting occurs.

This composite shows melt flux anomalies during sunny summertime conditions.

9/ No. 2: foehn winds. They’re responsible for classic C-shape pattern of melt anomalies w/ the highest melt in the lee of steep terrain. Foehn are esp important in MAM/SON when solar radiation is relatively less strong.

Fig: melt anoms during foehn conditions in MAM.

10/ How do we know about the relative importance of sunny/foehn conditions? Look at this 👀

DJF: lots melt with sunny, non-foehn conditions
MAM: melt w/ non-sunny foehn > sunny non-foehn
SON: -v anoms assoc w/ foehn and sunny alone, but +v when they act together (sunny foehn)

11/ What does this mean? Sunny = #1 melt driver in DJF, Foehn is #1 in MAM, you need both together to drive large melt in SON.

Fig: melt anomalies across seasons for combos of foehn/sunny

12/ 3rd place: #cloud. In DJF clouds stop SW reaching the surface, suppressing melt, but in MAM/SON they can have a thermal blanketing effect lg enough to raise temps and initiate / sustain melting, esp low-level liquid cloud.

Fig: melt anoms in DJF/MAM during cloudy conditions

13/ In DJF, you get lots of melt in sunny periods with thin, low LWP cloud. That can mean 1 of 2 things:

1. Hi-level cirrus with mostly ice (+foehn) (1st pic)
2. Low-level, low LWP cloud - thin enough to allow some SW thru, but also w/ strong LW warming effect (2nd pic)

14/ Although *absolute* melt production is low in cloudy conditions during MAM/SON, as #ClimateChange pushes temps higher, cloud might become more important, resulting in the ‘Greenlandification’ of Larsen C, w/ thin, low-level cloud increasing melt, esp in summer.

15/ Finally, the big pic: lg-scale circulation. SAM+, El Niño (ENSO-) and deep Amundsen Sea Low (ASL) all enhance melting by pushing air over the peninsula from the NW, resulting in foehn and reducing cloudiness.

In fact, SAM+ and foehn occurrence are +v’ly correl as shown here:

16/ Southerly barrier winds, SAM- and ENSO+ all suppress melting. You can see the summary figure of the top 3 melt enhancing and suppressing conditions here:

Fig: mean synoptic meteorology + Tmax anomalies and melt anomalies for each condition during DJF.

17/ Importantly, all melt drivers interact/overlap & can co-occur. To make things more complex, they also vary with the seasons (hence why it’s taken me 3 yrs to get this out!).

Here’s a schematic that summarises the most important modes of melting throughout the year.

18/ If you don’t wanna read the papers themselves (altho I would recommend ;) ) I got you!

👉 Check out my Carbon Brief guest post about the papers here: carbonbrief.org/guest-post-ran…

19/ Watch my videos about the papers here:

👉 part 1:
👉 part 2:

20/ Wanna use the data? It’s available on CEDA: catalogue.ceda.ac.uk/uuid/ce00cc319…

Tech spec:
Model: @MetOffice Unified Model (MetUM)
Resolution: 4 km
Time period: 1998-2017
Forcing: Era-Interim