Archivi categoria: WP1

Deliverable 1.3 – List of Indicators

Executive Summary

In task 1.3 a literature review was performed to individuate a set of environmental sustainability indicators (ESI) that allow the identification of the interconnections existing between the olive production and the surrounding environment. Meanwhile, thresholds, which will allow us to establish environmental sustainability references, were defined on the basis of expert’s evaluation after the review and stakeholder consultations in tasks 3.2 and 3.3. Indicators and thresholds will be used to compare different olives agro- ecosystems management options under current and future climate (WP4).

Read the complete deliverable 1.3

Deliverable 1.2 – Report on future climate scenarios

Introduction and objectives

Recent studies pointed out that since cultivated area of olive tree is limited to specific climatic niches, temperature and rainfall changes will put its cultivation at greater risk. Much evidence indicates that, during the last two millennia, the extension of its cultivated area changed and the climate was the main variable driving this process. Accordingly, climate change is expected to produce a large impact on olive tree distribution, where Southern Mediterranean region may become unsuitable for its cultivation because pf the joint effect of higher temperatures and reduced soil moisture while other regions, nowadays too cold, may become viable for olive tree cultivation following the warming trend.

Yield quality may be affected as well, since environment together with the cultivar is the factor with the highest impact on fruit and oil quality. Climate change is therefore raising serious concerns about possible variations caused by increasing temperatures and reduced rainfall, which are highly likely to be detrimental to fruit quality.

The aim of this task is to produce monthly and daily climate data (namely minimum and maximum temperature, rainfall and global radiation) under future RCP scenarios (4.5 and 8.5), for different time slices (2031-2040, 2041-2070, 2071-2100) at high spatial resolution (44 x 44 km) so as to respect OLIVECAN model protocol. Moreover, this task aimed at identifying different olive systems macro-categories on the basis of soil, climate and management parameters characterizing the Mediterranean basin on which simulating effects of improved management practices (WP4) and biotic stressors (WP2) through modelling (WP2) under an adaptive and mitigating perspective. According to the procedure agreed during the kick-off meeting in Florence, these task were summarized into a cluster classification of main climatological regions across the Mediterranean basin, as explained in the following paragraph.

Dataset

According to the procedure agreed during the kick-off meeting in Florence, the original dataset obtained by EOBS and aggregated to a spatial resolution of 10×10 km and described in deliverable D1.1, was further processed by clustering each grid cell to different homogeneous areas. Over these clustered areas, the outputs of Regional Circulation Models (RCMs) specifically developed for the Mediterranean region were downscaled, using a delta change approach. RCP45 and RCP85 scenarios were considered, which represent two possible radiative forcing values projected to 2100, +4.5 and +8.5 Wm-2 respectively. Three different RCMs were included in this process, namely CNRM-ALADIN52, CMCC-CCLM4-8-19 and MPI-ESM-LR.

Methodology

Clustering observed gridded dataset

The meteorological dataset spaces 10 by 10 km over the cultivated area of olive tree was further aggregated for speed up the Olivecan processing time. Monthly Climatological indices (average maximum and minimum temperatures, cumulated rainfall, minimum and maximum relative humidity and average wind speed) were firstly extracted starting from the daily meteorological series of each cell. These data were further processed to get the maximum of monthly maximum temperatures, the minimum of monthly minimum temperatures, the maximum of monthly-cumulated rainfall, the yearly-cumulated rainfall, radiation, and the maximum monthly wind speed. This matrix, consisting of 4776 grids by 9 dimensions was processed using The High Dimensional Discriminant Analysis and Data Clustering (HDDC) that is suitable unsupervised classification with high dimensional data. Further, Since HDDC is a model-based clustering method, the Bayesian information criterion (BIC) may be used to select the number of clusters K to keep. HDDC, implemented in the R package HDclassif provides a simple way to do this: it displays the BIC value for each clustering result for different number of classes and select the model, which maximizes it. Accordingly, the selected matrix was clusterised under HDDC algorithm in order to detect the model providing the best BIC and the associated number of cluster.

RCMs Temperature and Precipitation

Monthly climate data of minimum and maximum temperature, cumulated rainfall and global radiation (Tmin, Tmax, R and Rad ) for the baseline (1981-2005) and future time slices 2011-2040, 2041-2070, 2071-2100 as simulated by RCM CNRM-ALADIN52, MPI-ESM-LR and CMCC-CCLM4-8-19 for RCP 45 and 85, at a spatial resolution of 0.44°x0.44° (https://www.medcordex.eu/) were downscaled over observed data clustered according to 4.1. Specifically, according to delta change procedure, the monthly average differences between the baseline and future time-slices for Tmin, Tmax, R and Rad were added to the observed daily dataset of each cluster in order to derive relevant dataset to be used as input variable for OLIVECAN for future periods.

Results and discussion

Cluster analysis

The analysis of BIC criterion indicated that 19 clusters significantly different each other as outlined in figure 1 and summarised in table 1.The daily data of grid points were aggregated for each cluster calculating the daily average temperatures, rainfall, relative humidity and global radiation.

Figure 1. Geographical distribution of 19 climatic areas across the olive tree cultivated area.
Figure 1. Geographical distribution of 19 climatic areas across the olive tree cultivated area.

 

Table 1. Average values of each index reported for the 19 identified cluster. X and Y represent the centroid of the cluster

Future Temperature and Precipitation

In average, the different RCMs selected for this project showed the same trend for changes in temperature, rainfall and radiation pattern, but with a different magnitude. Change in temperature progressively increased from 2011-2040 to 2071-2100 with the highest increased observed for rcp85 scenario, with a not clear seasonal pattern. Not a clear difference was observed amongst the RCMs used in this project. On a spatial basis, Southern France, Spain and Italy were the most affected by increased temperature especially in summer time.

Rainfall changes exhibited a clear seasonal pattern in all considered RCMs, where winter and fall seasons were the less affected or showed even an increased trend, while in any case, Spring and Summer seasons showed the highest decrease in cumulated rainfall. This trend has a different magnitude, where RCP85 resulted into a more severe impact with respect toRCP85.

RCM CNRM-ALADIN52 predicted a lower effect of both emission scenarios on rainfall, while the highest impact was recorded for CMCC-CCLM4-8-19.

Conclusion

The task objectives were completely achieved and the resulting dataset is available to feed OLIVECAN model for both present and next future under 2 emission scenarios and 3 timeslices.

Read the complete deliverable 1.2

Deliverable 1.1 – Report on data collection and relevant metadata

Executive summary

Purpose

The aim of this report is to provide all the necessary data required for a full calibration of the OLIVECAN model in a wide range of olive tree varieties, management practices and environmental conditions across Mediterranean region.
Methods

All partners provided all the necessary data to spatially identify current olive tree distribution across Mediterranean countries through pre-existing database. Partners provided data regarding main soil and climatic characteristics. Additionally, they included all the available data regarding the main management practices that are implemented in Cyprus, Italy, Spain and Greece. Data were qualitatively checked to produce homogeneous and reliable data. Furthermore, a standardized template was prepared regarding the description of experimental sites and the data needed (meteorological and climate data) for the calibration of OLIVE-CAN model and to reproduce suitable olive tree distribution using a statistical model (Random Forest) under current and future climate.

Results

In this report olive tree distribution, soil characteristics (soil texture, organic carbon, pHorganic Carbon, pH, water storage capacity, soil depth, cation exchange capacity of the soil and the clay fraction, total exchangeable nutrients, lime and gypsum contents, sodium exchange percentage, salinity, textural class and granulometry), meteorological data (temperature, RH, percipitation) were down scaled to the to the 10×10 km grid covering the olive tree distribution, by using the overlay tools in a GIS environment. Management practices differs among countries and OLIVE-MIRACLE recorded average information collected from Italy, Spain, Greece and Cyprus. Finally the description of the experimental sites has been set and the climatic and meteorological data were collected. The data regarding the sites were substantially different among countries and are provided as metadata tables in the appendixes of the current report. spatial information (i.e. olive tree distribution, soil, climate and olive tree management) at a courser resolution (10×10 km).

Conclusions

The deliverable provided all the necessary dataset required for the OLIVECAN model spatially harmonized and integrated within a database. Through this deliverable we are able to identify and characterize representative olive systems macro-categories across the northern Mediterranean basin, aggregating relevant spatial information for the at a courser resolution (10×10 km) (D1.2).

Read the complete deliverable 1.1