Response in atmospheric circulation and sources of Greenland precipitation to glacial boundary conditions

Abstract

The response in northern hemisphere atmospheric circulation and the resulting changes in moisture sources for Greenland precipitation to glacial boundary conditions are studied in NCAR’s CCM3 atmospheric general circulation model fitted with a moisture tracking functionality. We employ both the CLIMAP SST reconstruction and a modification thereto with reconstructions of glacial ice sheets and land masks. The individual components of the boundary conditions are added first one at a time and, finally, together. These steps show the atmospheric circulation to respond approximately linearly to the boundary condition changes, and the full glacial change may thus be decomposed into contributions from SST and topography changes, respectively. We find that using the CLIMAP SST reconstruction leads to a shift from Atlantic toward Pacific source regions not found with the modified reconstruction having cooler tropics and less sea ice. The occurrence of such a shift depends chiefly on the SST reconstruction and not on the existence of the large northern hemisphere glacial ice sheets. The influence of these circulation changes on important factors for ice core interpretation such as precipitation seasonality, condensation temperatures and source temperatures are assessed.

Appendix: Moisture tracking

The moisture tracking employed here in the CCM3 resembles that described by Joussaume et al. (1986). The moisture field is split into contributions from a number of regions into which the Earth’s surface has been divided (see Fig. 3a):

$$Q = \sum_{m}Q_{m},$$

(8)

where the sum is taken over all the regions. These moisture classes are treated as tracers following the same advective, diffusive and adjustment processes as the model’s moisture field. The above sum of the tracers is at each time step and each grid point forced to equal the model’s total moisture (as also done by, e.g., Noone and Simmonds 2002) by a common scaling of each tracer.

The dynamical advection and diffusion of the tracers is accomplished by the distributed model’s built-in infrastructure for tracers (for details, see Kiehl et al. 1996). This uses a semi-Lagrangian transport (SLT) scheme for both the model’s moisture field and for tracers. In this manner, the transport of the individual moisture tracers parallels that of the total moisture field. Due to the fact that the SLT is non-conservative, a fixer is applied to the fields at each time step to assure global conservation (Kiehl et al. 1996). However, if one tracer is adjusted in a grid point, subsequent enforcement of Eq. (8) can lead to a net transformation of one tracer into another. Nevertheless, we have checked the consistency between the total amount of evaporated and precipitated tracer (“what goes up must come down”) and found that the majority of the tracers are conserved within 1% and all are conserved within 3%.

Surface exchange As in Joussaume et al. (1986) the evaporative flux, E, is considered at each grid point at each time step. For a grid point belonging to the mth source region, if the flux is positive (upward),

$$E_{m} = E $$

(9)

and it contributes to the tracer Q _m in the lowest atmospheric layer. If the evaporative flux is negative (downward), condensation of moisture onto the surface occurs. In this case, all tracers are affected by contributing to the downward flux in ratios corresponding to their share of the total moisture in the lowest atmospheric level:

$$E_{r} = E \frac{Q_{r}(k^{\max})}{Q(k^{\max})},$$

(10)

for all tracer indices, r. Here k ^max = 18 denotes the lowest atmospheric level.

Adjustment physics The model’s adjustment physics routine (aphys.F and subroutines) calculates convective mass fluxes, condensation, precipitation, re-evaporation of falling rain etc. The changes in the set of tracers constituting the total moisture field are calculated by dividing the changes into those due to convective redistribution and those due to removal by condensation. In the following we will look at one atmospheric column at a time and latitude-longitude indices will thus be omitted. Consider a change in the moisture content at level k, δQ(k) = Q′(k) −Q(k), from Q(k) to Q′(k) due to the adjustment physics. We now search the column for layers with positive and negative changes, respectively. Summing over these gives us the total positive and negative changes:

$$\delta Q^+ = \sum_{\delta Q(k)\ge 0} \delta Q(k)$$

(11)

$$\delta Q^- = \sum_{\delta Q(k) < 0} \delta Q(k). $$

(12)

These changes are then split into contributions from convective redistribution and condensation. In “positive” layers, all change is assumed to be due to convection (i.e., we disregard the possibility of condensation from a layer followed by convective input leading to a net positive change):

$$ \delta Q^{cnv}(k) = \delta Q(k), \quad \delta Q^{cnd}(k) = 0, \quad (\hbox{pos}) $$

(13)

where the cnv and cnd superscripts refer to convective and condensation parts, respectively. In “negative” layers, the convective parts contribute to the total positive change in ratios corresponding to their share of the total negative change:

$$\delta Q^{cnv}(k) = -\frac{\delta Q(k)}{ \delta Q^-} \delta Q^+, \quad (\hbox{neg}) $$

(14)

while the rest is due to condensation:

$$\delta Q^{cnd}(k) = \delta Q(k)-\delta Q^{cnv}(k) = \delta Q(k)\left (1+ \frac{\delta Q^+} {\delta Q^-}\right ). \quad (\hbox{neg}) $$

(15)

These δQ ^cnd(k) from negative layers constitute the precipitation from the column. Note that −1 ≤ δQ ⁺/δQ ⁻ ≤ 0, since δQ ⁻ ≤ 0 and the adjustment physics can only redistribute or remove water, such that |δQ ⁻| ≥ δQ ⁺.

Turning to the tracer budgets, we consider the mth tracer, Q _m . In negative layers, the change due to convection and condensation are calculated from the ratio of the tracer at that level, Q _m (k), to the total moisture;

$$\delta Q_{m}^{cnv}(k) = \delta Q^{cnv}(k)\frac{Q_{m}(k)}{Q(k)}, \quad \delta Q_{m}^{cnd}(k) = \delta Q^{cnd}(k)\frac{Q_{m}(k)}{Q(k)}.\quad (\hbox{neg}) $$

(16)

The total of the negative convective changes,

$$\delta Q_{m}^{cnv-} =\sum_{\delta Q(k) < 0} \delta Q_{m}^{cnv}(k) , $$

(17)

is distributed among the positive-change layers in ratios corresponding to the total positive change:

$$\delta Q_{m}^{cnv}(k) = -(\delta Q_{m}^{cnv-}){\delta Q^{cnv}(k)} {\delta Q^+}. \quad (\hbox{pos}) $$

(18)

Since ∑ _m Q _m (k) = Q(k) is enforced at every time step, the convective and condensation changes tracer-by-tracer can be shown to add up to the corresponding changes in the full moisture field.