TCN	Temporal Convolutional Network
SNA	Spatial Neighborhood Attention
CSTA	Cross-Scale Trend Attention
MLP	Multi-Layer Perceptron
SSM	State-Space Model
VDI	Ventilation Demand Index
${\overline{T}}_i$	Average temperature of the i-th spatial block
$V_i$	Temperature variance of the i-th spatial block
$G_{X,i}$	Temperature gradient of the i-th spatial block in the row (X) direction
$G_{Y,i}$	Temperature gradient of the i-th spatial block in the column (Y) direction
$G_{Z,i}$	Temperature gradient of the i-th spatial block in the layer (Z) direction
N	Neighborhood Set consisting of the Target Location and Spatial Neighbors
$B_c$	Target Location
$B_k$	The k-th Spatial Neighbor
$X_{\mathrm{in}}$	the input sequence
$T_i(t)$	the thermal state features of a spatial block
$E(t)$	Meteorological information at time step t
$S_i(t)$	Spatial feature representation generated by the SNA module at time step t
$H_{\mathrm{TCN}}$	Feature representation extracted by the Hierarchical TCN branch
$H_{\mathrm{Mamba}}$	Feature representation extracted by the Mamba branch
$H_f$	Fused feature representation of the TCN and Mamba branches
Z	Multi-scale trend embedding
$g_R$	Adaptive fusion weights for recent-term trends
$g_S$	Adaptive fusion weights for short-term trends
$g_L$	Adaptive fusion weights for long-term trends
$H_{\mathrm{Trend}}$	Fused trend representation
q	Heat flux density vector
k	Thermal conductivity
∇T	Three-dimensional temperature gradient
$P_i^t$	Recommended ventilation-priority level
L	Length of the historical input sequence
t	Time index