Onshore transport - needed for autocyclicity

[EmiliaJarochowska]

I should have identified this earlier, but it slipped my attention - sorry! For the emergence of autocyclicity in the absence of external drivers, it seems that onshore transport is required. Please see Cases 2 and 3 in Xi & Burgess (2022). They write:

As in previous models (Burgess et al., 2001), onshore prevailing winds and currents are assumed to produce a time-averaged net onshore advective sediment flux.

In top (plane) view, this is the regularity obtained:

Our version of CarboCAT doesn't have this feature, I wrote to Peter asking for that specific part of the code. This issue needs discussion between us first before anyone works on it.

[Burgesski]

I have had a look at the version of CarboCAT that Haiwei modified and used to produce these autocycle models. It's just a smidgeon convoluted ;-) so apologies for that, but I think actually it can be reasonably easily unravelled to understand the key elements in the sediment transport formulation that produces the shoreline/island progradation that generates the shallowing-upward autocycles.

The key functions are the various versions of the sediment transport calculation function and they seem to be calculateTransportStraight, calculateTransportPlatformTopShoaling, and calculateTransportPlatformTopShoalingWaveRefraction

I know, I know, multiple versions of essentially the same function calculate the sediment transport in slightly different ways is slightly sub-optimal coding practice ;-) but the good news is that the most important element in the calculation is the same in each function.

In each case a volume (thickness actually) of sediment is entrained, based on a parameter indicating the proportion of produced sediment to be transported, as per standard CarboCAT practice. Looking in calculateTRansportStraight.m in lines 133 - 150 we can see that this sediment is then moved across the platform top following a defined vector, for example straight up the y-axis of the model grid, so long as the following while loop condition is met

while foundLocalHigh == false && endOfFlowVectorReached == false && flowLength > 1

flowLength and endOfFlowVectorReached just deal with no transport path defined and end of defined path reached, for example at the end of the grid, so the foundLocalHigh flag is the element we are most interested in, and this is defined via

if grad > glob.depoGradThreshold
foundLocalHigh = true;
DepobyLocalHighs(flowVect(2,j), flowVect(1,j))=1;
end

which sets foundLocalHigh to true as the while loop terminating flag when the gradient into the next cell along the transport path exceeds a defined threshold.

So in summary, sediment is entrained from each producing point on the grid, and then transported across the grid until reaching a point where there is a topographic gradient to the next grid cell in the transport path, from deeper water into shallower water, that exceeds a defined threshold value, at which point the sediment is deposited.

If you can implement something like this in CarboKitten I would anticipate shoreline/island progradation behaviour should result, but I'm not sure how straight forward this will be to do given the diffusion approach adopted in CarboKitten? Perhaps we need to discuss this further??

cheers, P.