Impacts of land-use and land-cover changes on temperature-related mortality

Abstract

Background: Land-use and land-cover change (LULCC) can substantially affect climate through biogeochemical and biogeophysical effects. Here, we examine the future temperature-mortality impact for two contrasting LULCC scenarios in a background climate of low greenhouse gas concentrations. The first LULCC scenario implies a globally sustainable land use and socioeconomic development (sustainability). In the second LULCC scenario, sustainability is implemented only in the Organisation for Economic Cooperation and Development countries (inequality).

Methods: Using the Multi-Country Multi-City (MCC) dataset on mortality from 823 locations in 52 countries and territories, we estimated the temperature-mortality exposure-response functions (ERFs). The LULCC and noLULCC scenarios were implemented in three fully coupled Earth system models (ESMs): Community Earth System Model, Max Planck Institute Earth System Model, and European Consortium Earth System Model. Next, using temperature from the ESMs' simulations and the estimated location-specific ERFs, we assessed the temperature-related impact on mortality for the LULCC and noLULCC scenarios around the mid and end century.

Results: Under sustainability, the multimodel mean changes in excess mortality range from -1.1 to +0.6 percentage points by 2050-2059 across all locations and from -1.4 to +0.5 percentage points by 2090-2099. Under inequality, these vary from -0.7 to +0.9 percentage points by 2050-2059 and from -1.3 to +2 percentage points by 2090-2099.

Conclusions: While an unequal socioeconomic development and unsustainable land use could increase the burden of heat-related mortality in most regions, globally sustainable land use has the potential to reduce it in some locations. However, the total (cold and heat) impact on mortality is very location specific and strongly depends on the underlying climate change scenario due to nonlinearity in the temperature-mortality relationship.

Keywords: Deforestation; Land-use and land-cover change; Mortality; Sustainable land use; Temperature.

Figure 1.

Schematic presentation of the two LULCC scenarios (inequality and sustainability) and modeling interface. The scenarios are described in detail in Humpenöder et al.

Figure 2.

Average daily mean temperature (°C) at the 823 MCC locations across 52 countries and territories in five inhabited continents used in this study. The daily mean temperature is computed using surface observations from the MCC data and averaged over the location-specific time periods shown in Table S1; http://links.lww.com/EE/A302.

Figure 3.

Change in near-surface mean air temperature (°K) by 2069–2099. The noLULCC scenario is compared to the historical control (1980–2014). The sustainability and inequality scenarios are compared to the noLULCC scenario (2069–2099). The dots indicate the consistency in terms of the sign of temperature response across the three individual ensemble members of the ESMs.

Figure 4.

Multimodel mean changes in the fraction of total (cold and heat) excess mortality (percentage points) in 2050–2059 (left panel) and 2090–2099 (right panel) under noLULCC relative to a historical period of 1980–1989. The blue circles, red triangles, and purple rectangles, respectively, represent cold-related, heat-related, and total EM-derived estimates for 823 MCC locations across 14 geographic regions. The world regions are ordered by latitude on the Y-axis. The numbers following the region names on Y-axis show the number of MCC locations. Uncertainty in the empirical distribution of EM (i.e., 2.5th and 97.5th percentiles) are estimated using Monte Carlo simulations and shown in Table S2; http://links.lww.com/EE/A302.

Figure 5.

Same as Figure 4 but for the sustainability scenario relative to the noLULCC scenario by 2050–2059 (left panel) and 2090–2099 (right panel). The X-axis ranges in this figure differ from those in Figure 4.

Figure 6.

Same as Figure 4 but for the inequality scenario relative to the noLULCC scenario by 2050–2059 (left panel) and 2090–2099 (right panel). The X-axis ranges in this figure differ from those in Figures 4 and 5.

Figure 7.

Regionally aggregated multimodel mean change in the fraction of total excess temperature-related mortality fractions (percentage points) in 2050–2059 (left panel) and 2090–2099 (right panel). The red, blue, and green circles show the mortality impacts for the noLULCC, inequality, and sustainability scenarios, respectively, grouped by 14 geographic regions. The numbers following the region names on the Y-axis show the number of MCC locations. The regions with a small number of locations (e.g., S-Asia and Caribbean) are underrepresented.