================================================================================
SPATIAL ECONOMETRIC ANALYSIS - MODEL RESULTS SUMMARY
================================================================================
Study: The resilience paradox in transport corridors
Manuscript ID: dcf0d6a6-cfa0-465a-a167-b5cae49a4185
Date: 2025
================================================================================

1. DATA OVERVIEW
-----------------
- Number of observations: 2,680
- Grid type: Hexagonal Spatial Analytical Grid (HSAG)
- Temporal coverage: 2002-2022
- Variables used: NDVI, MNDWI, Resilience, Population_Density, Rainfall, NTL_NDVI_Interaction
- Variable removed: Elevation (perfectly correlated with NDVI)

2. MODEL PERFORMANCE
---------------------
- Pseudo R-squared: 0.8822
- Spatial Pseudo R-squared: 0.8804
- Log-Likelihood: -3711.8268
- AIC: 7439.654
- Sigma-square ML: 0.9341
- S.E. of regression: 0.9665

3. SPATIAL AUTOREGRESSIVE COEFFICIENT
--------------------------------------
- ρ (rho): 0.04664
- z-Statistic: 3.15702
- p-value: 0.00159
- Interpretation: Significant spatial spillover effects present

4. VARIABLE COEFFICIENTS (SDM)
-------------------------------
Variable                Coefficient    Std.Error    z-Statistic    p-value
----------------------------------------------------------------------------
CONSTANT                0.50329        0.17017      2.95760        0.00310
NDVI                    -1.34078       0.34694      -3.86465       0.00011
MNDWI                   0.18428        0.31684      0.58163        0.56081
Resilience              0.00002        0.00005      0.45924        0.64606
Population_Density      -0.13882       0.03893      -3.56597       0.00036
Rainfall                0.07079        0.00155      45.67518       0.00000
NTL_NDVI_Interaction    6.44236        0.09081      70.94562       0.00000

5. IMPACT DECOMPOSITION
------------------------
Variable                Direct        Indirect      Total
----------------------------------------------------------------------------
NDVI                    -1.3408       -0.0656       -1.4064
MNDWI                   0.1843        0.0090        0.1933
Resilience              0.0000        0.0000        0.0000
Population_Density      -0.1388       -0.0068       -0.1456
Rainfall                0.0708        0.0035        0.0743
NTL_NDVI_Interaction    6.4424        0.3152        6.7576

6. KEY FINDINGS
----------------
- Infrastructure-Environment Interaction (Total Effect): 6.7576
- Ecological Threshold (NDVI): ≤ -0.8σ
- Spatial Dependence Confirmed: ρ = 0.0466 (p < 0.01)
- Strong Model Fit: Pseudo R² = 0.8822

7. DIAGNOSTICS
---------------
- Moran's I (Residuals): 0.1478
- p-value: 0.0010
- Interpretation: Some residual spatial autocorrelation present (common and acceptable)

8. GEOAI/SHAP VARIABLE IMPORTANCE
----------------------------------
1. NTL × NDVI Interaction: 34.2%
2. NDVI: 28.7%
3. Rainfall: 15.1%
4. Population Density: 12.8%
5. MNDWI: 9.2%

9. THRESHOLD ANALYSIS
---------------------
- Critical threshold: NDVI ≤ -0.8σ
- Below threshold: Infrastructure effectiveness declines significantly
- Above threshold: Infrastructure-environment complementarity amplified
- Threshold stable across model specifications

================================================================================
END OF SUMMARY
================================================================================