Evaluating the spatial coherence of composite indexes of land degradation at small administrative units

Abstract

In light of the ‘zero-net land degradation strategy’, the United Nations Convention to Combat Desertification (UNCCD) guidelines for policy-relevant indicators of land quality, have stressed the importance of comparable indexes' informative value across spatial scales. However, the proliferation of composite indexes of land sensitivity, vulnerability, or exposure to desertification risk, has been occasionally grounded on a systematization of theoretical concepts – both in advanced economies and in emerging countries. Rationalization of the operational approaches to land degradation assessment implies a (spatially explicit) comparison and verification of their outcomes at different (policy-relevant) spatial scales. To answer this issue, we implemented a statistical approach evaluating the coherence and reliability of four composite indexes of land degradation (ES3, ES4, LVI, DSI), with a practical application for Italy, a Mediterranean country classified as ‘affected’ by UNCCD. Using (municipal) administrative units as the elementary analysis' domain, sequential econometric techniques were used to test the internal (spatially explicit) coherence of the four composite indexes, and the role of information redundancy, for calibration of local-scale land degradation strategies. The outcomes of spatial regressions documented the potential of the four indexes in a short-term assessment of land vulnerability, confirming that such indexes provided a coherent ranking of Italian municipalities based on their intrinsic degree of land degradation. Principal Component Analysis documented the superior informative power of the ES4 index and confirms the appropriateness of an extensive use of the Environmentally Sensitive Area Index (ESAI) at small spatial scales. The proposed framework supports the creation of an index dashboard to monitoring complex environmental phenomena.