Land cover changes and climate extremes critically reshape ecosystem services (ES). Yet the response of ES to these drivers as indicators of ecosystem degradation or restoration remains understudied in semi-arid regions. This research evaluates and compares fluctuations in ecosystem water balance residual (eWBR), carbon storage (CS), and carbon sequestration (C.Seq) in response to land conversions and climate extremes in the Lake Chad Basin (LCB) and Aral Sea Basin (ASB), with implications for achieving Sustainable Development Goals (SDGs). The objectives are to: (i) simulate land cover changes and associated ES parameters; (ii) identify key predictors of ES; and (iii) quantify the magnitude and spatial extent of ES responses to land conversions and climatic extremes. We integrated cellular automata, Markov chains, and multi-layer perceptron models to simulate land cover change and validated the results using the Kappa coefficient. The InVEST model was calibrated with localized parameters for arid/semi-arid ES to simulate CS, C.Seq, and the eWBR. These ES were analyzed using Theil-Sen slopes for trends, ordinary least squares regression for identifying predictors, pixel correlations, and advanced geostatistics quantified the spatial variability and response to land cover change and climate extremes. Key findings revealed baseline degradation in both basins, with C.Seq (coefficients 1.858 and 0.00 gC/m2/y-1) and CS (coefficients -0.025 and -0.002 gC/m2/y-1) in LCB and ASB. In LCB, temperature and NDVI predicted tradeoffs between eWBR and CS (coefficients -0.210 and -0.011), while temperature and precipitation drove synergies (coefficients 0.033 and 0.045) in ASB. Land conversions had divergent impacts: The conversion of shrublands to cropland degraded CS by -0.017 gC/m2/y-1, while cropland conversion to shrubland improves CS and C.Seq by 0.23 and 0.01gC/m2/y-1 in the LCB. Contrastingly, in the ASB, the conversion of water to bare land decreased CS by -0.09 gC/m2/y-1, and the conversion of bare land to cropland improved CS and C.Seq by 0.090.51 gC/m2/y-1 each. In the LCB, increased precipitation and cooling restored CS (0.51-0.54 gC/m2/y-1) but reduced C.Seq (- 0.04 to - 0.97 gC/m2/y-1). In the ASB, a decrease in precipitation restored CS by 0.39 gC/m2/y-1 but degraded eWBR by -0.30 mm/y-1, while warming improved CS (0.31 gC/m2/y-1) and C.Seq (-0.01 gC/m2/y-1). Although limited to remotely sensed and modeled data, these findings emphasize the need for functional land management, temperature and precipitation-resilient strategies, and transboundary cooperation to support ES restoration and minimize degradation, informing both policy and management practices aimed at achieving SDGs.