correct citet

docs/src/examples/glacialcycle.jl

@@ -8,7 +8,7 @@ We now want to provide an example that presents:
 - a more elaborate load that evolves over time
 - changes in the sea-level
 
-For this we run a glacial cycle of Antarctica with lithospheric thickness and upper-mantle viscosity from @wiens-seismic-2022 and the ice thickness history from @peltier-comment-2018. We start by generating a [`ComputationDomain`](@ref) with intermediate resolution for the sake of the example and load the heterogeneous lithospheric from @pan-influence-2022 thanks to the convenience of [`load_dataset`](@ref):
+For this we run a glacial cycle of Antarctica with lithospheric thickness and upper-mantle viscosity from @wiens-seismic-2022 and the ice thickness history from [peltier-comment-2018](@citet). We start by generating a [`ComputationDomain`](@ref) with intermediate resolution for the sake of the example and load the heterogeneous lithospheric from [pan-influence-2022](@citet) thanks to the convenience of [`load_dataset`](@ref):
 =#
 
 using CairoMakie, FastIsostasy
@@ -32,7 +32,7 @@ end
 nicer_heatmap(Tlitho)
 
 #=
-In a similar way, we can load the log-viscosity field from @pan-influence-2022 and plot it at about 300 km depth
+In a similar way, we can load the log-viscosity field from [pan-influence-2022](@citet) and plot it at about 300 km depth
 =#
 
 (_, _, _), _, logeta_itp = load_logvisc_pan2022()
@@ -95,5 +95,5 @@ Colorbar(fig[1, 4], height = Relative(0.6); opts...)
 fig
 
 #=
-The displayed fields are displacement anomalies w.r.t. to the last interglacial, defined as the reference for the ice thickness anomalies. In @swierczek2024fastisostasy these computations are performed on a finer grid, with an interactive sea level, and show great agreement with a 3D GIA model that runs between 10,000-100,000 slower (however at with advantage of obtaining a global and richer output). You can find this more comprehensive example in the `/publication_v1.0` folder of the GitHub repository.
+The displayed fields are displacement anomalies w.r.t. to the last interglacial, defined as the reference for the ice thickness anomalies. In [swierczek2024fastisostasy](@citet) these computations are performed on a finer grid, with an interactive sea level, and show great agreement with a 3D GIA model that runs between 10,000-100,000 slower (however at with advantage of obtaining a global and richer output). You can find this more comprehensive example in the `/publication_v1.0` folder of the GitHub repository.
 =#

docs/src/index.md

@@ -4,9 +4,9 @@
 
 ![GlacialCycle](assets/isl-ice6g-N=350.gif)
 
-FastIsostasy is a friendly and flexible model that regionally computes the glacial isostatic adjustment (GIA) with laterally-variable mantle viscosity and lithospheric thickness. It is described in @swierczek2024fastisostasy and is mainly adressed to ice-sheet modellers who seek for (1) a good representation of solid-Earth mechanics at virtually zero computational cost, (2) an approximation of the sea-level equation and (3) ready-to-use inversion tools to calibrate the model parameters to data. The simple interface of FastIsostasy allows to flexibly solve GIA problems within few lines of code. The code is distributed under GNU general public license v3 and was succesfully benchmarked against analytical, 1D GIA and 3D GIA model solutions.
+FastIsostasy is a friendly and flexible model that regionally computes the glacial isostatic adjustment (GIA) with laterally-variable mantle viscosity and lithospheric thickness. It is described in [swierczek2024fastisostasy](@citet) and is mainly adressed to ice-sheet modellers who seek for (1) a good representation of solid-Earth mechanics at virtually zero computational cost, (2) an approximation of the sea-level equation and (3) ready-to-use inversion tools to calibrate the model parameters to data. The simple interface of FastIsostasy allows to flexibly solve GIA problems within few lines of code. The code is distributed under GNU general public license v3 and was succesfully benchmarked against analytical, 1D GIA and 3D GIA model solutions.
 
-FastIsostasy relies on a hybrid Fourier/finite-difference collocation of the problem introduced in [cathles-viscosity-1975](@cite) and solved in [lingle-numerical-1985](@cite), [bueler-fast-2007](@cite). Thanks to a simplification of the full problem from 3D to 2D space and the use of [optimized software packages](@ref Juliaecosystem), running kiloyears of regional GIA with $$\Delta x = \Delta y = 45 \, \mathrm{km}$$ is a matter of seconds on a single CPU. For high resolution runs, the user can switch to GPU usage with minimal syntax change and enjoy the advantage of parallelization without requiring an HPC cluster. For GIA "purists", this package is likely to miss interesting processes but we belive that its ridiculous run-time can help to fast-prototype a problem before transfering it to a more comprehensive model.
+Based on the work of [cathles-viscosity-1975](@citet) and [lingle-numerical-1985](@citet), an efficient way of solving for the vertical displacement was proposed by [bueler-fast-2007](@cite). FastIsostasy generalises this approach by relying on a hybrid Fourier/finite-difference collocation. Thanks to a simplification of the full problem from 3D to 2D space and the use of [optimized software packages](@ref Juliaecosystem), running kiloyears of regional GIA with $$\Delta x = \Delta y = 45 \, \mathrm{km}$$ is a matter of seconds on a single CPU. For high resolution runs, the user can switch to GPU usage with minimal syntax change and enjoy the advantage of parallelization without requiring an HPC cluster. For GIA "purists", this package is likely to miss interesting processes but we belive that its ridiculous run-time can help to fast-prototype a problem before transfering it to a more comprehensive model.
 
 !!! tip "Star us on GitHub!"
     If you have found this library useful, please consider starring it on [GitHub](https://github.com/JanJereczek/FastIsostasy.jl). This gives us a lower bound of the satisfied user count.

docs/src/introGIA.md

@@ -1,6 +1,6 @@
 # Quick intro to GIA
 
-Glacial isostatic adjustment (GIA) denotes the crustal displacement that results from changes in the ice, liquid water and sediment columns, as well as associated changes in Earth's gravity and rotation axis, ultimately impacting the sea level. The magnitude and time scale of the deformational response depends on the applied load and on solid-Earth parameters, i.e. the mantle viscosity, the lithosphere thickness and their respective density. These parameters display a radial and sometimes also a lateral variability, further jointly denoted by parameter "heterogeneity". For further details, please refer to [wiens-seismic-2022](@cite) and [ivins-antarctic-2022](@cite).
+Glacial isostatic adjustment (GIA) denotes the crustal displacement that results from changes in the ice, liquid water and sediment columns, as well as associated changes in Earth's gravity and rotation axis, ultimately impacting the sea level. The magnitude and time scale of the deformational response depends on the applied load and on solid-Earth parameters, i.e. the mantle viscosity, the lithosphere thickness and their respective density. These parameters display a radial and sometimes also a lateral variability, further jointly denoted by parameter "heterogeneity". For further details, please refer to [wiens-seismic-2022](@citet) and [ivins-antarctic-2022](@citet).
 
 ## Why do we care?
 
@@ -15,13 +15,13 @@ GIA models present a wide range of complexity, which can only be briefly mention
 2. The heterogeneity of the lithospheric thickness and upper-mantle viscosity cannot be represented.
 3. Changes in sea-surface height due to changes in mass repartition are ignored.
 
-On the higher end of the complexity spectrum, we find the 3D GIA models which address all the limitations of low-complexity models but are expensive to run, more tedious to couple to an ice-sheet model and generally lack a well-documented and open-source code base. Due to these drawbacks, they do not represent a standard tool within the ice-sheet modelling community. Although, they are becoming increasingly used, as for instance in [gomez-coupled-2018](@cite) and [van-calcar-simulation-2023](@cite), we believe that the expense of 3D GIA models can be avoided while still addressing the aforementioned limitations of simplistic models. Models specifically designed for ice-sheet modelling, such as [bueler-fast-2007](@cite) and [coulon-contrasting-2021](@cite), have shown first improvements in closing the gap between simplistic and expensive models. FastIsostasy continues this work by generalizing both of these contributions into one
+On the higher end of the complexity spectrum, we find the 3D GIA models which address all the limitations of low-complexity models but are expensive to run, more tedious to couple to an ice-sheet model and generally lack a well-documented and open-source code base. Due to these drawbacks, they do not represent a standard tool within the ice-sheet modelling community. Although, they are becoming increasingly used, as for instance in [gomez-coupled-2018](@citet) and [van-calcar-simulation-2023](@citet), we believe that the expense of 3D GIA models can be avoided while still addressing the aforementioned limitations of simplistic models. Models specifically designed for ice-sheet modelling, such as [bueler-fast-2007](@citet) and [coulon-contrasting-2021](@citet), have shown first improvements in closing the gap between simplistic and expensive models. FastIsostasy continues this work by generalizing both of these contributions into one.
 
 We here omit to speak about other GIA models, since they lack the representation of heterogeneous solid-Earth parameters.
 
 ## FastIsosatsy.jl in the model hierarchy
 
-FastIsostasy is capable of regionally reproducing the behaviour of a 3D GIA model at a computational cost that is reduced by 3 to 5 orders of magnitude. It relies on LV-ELVA, a generalisation of [bueler-fast-2007, coulon-contrasting-2021](@cite), and on the Regional Sea-Level Model (ReSeLeM).
+FastIsostasy is capable of regionally reproducing the behaviour of a 3D GIA model at a computational cost that is reduced by 3 to 5 orders of magnitude. It relies on LV-ELVA, a generalisation of [bueler-fast-2007, coulon-contrasting-2021](@citet), and on the Regional Sea-Level Model (ReSeLeM).
 
 FastIsostasy heavily relies on the Fast-Fourier Transform (FFT), as (1) its central PDE is solved by applying a Fourier collocation scheme and (2) important diagnostic fields are computed by matrix convolutions which can famously be accelerated by the use of FFT. FFT therefore inspired the name "FastIsostasy", along with a [GitHub repository](https://github.com/bueler/fast-earth) that eased the first steps of this package. The use of a performant language such as julia, as well as supporting performance-relevant computations on GPU allows FastIsostasy to live up to the expectations of low computation time.