Forecasting Vegetation Condition with a Bayesian Auto-regressive Distributed Lags (BARDL) Model

Abstract

Abstract. Droughts form a large part of climate/weather-related disasters reported globally. In Africa, pastoralists living in the Arid and Semi-Arid Lands (ASALs) are the worse affected. Prolonged dry spells that cause vegetation stress in these regions have resulted in the loss of income and livelihoods. To curb this, global initiatives like the Paris Agreement and the United Nations recognised the need to establish Early Warning Systems (EWS) to save lives and livelihoods. Existing EWS use a combination of Satellite Earth Observation (EO) based biophysical indicators like the Vegetation Condition Index (VCI) and socio-economic factors to measure and monitor droughts. Most of these EWS rely on expert knowledge in estimating upcoming drought conditions without using forecast models. Recent research has shown that the use of robust algorithms like Auto-Regression, Gaussian Processes and Artificial Neural Networks can provide very skilled models for forecasting vegetation condition at short to medium range lead times. However, to enable preparedness for early action, forecasts with a longer lead time are needed. The objective of this research work is to develop models that forecast vegetation conditions at longer lead times on the premise that vegetation condition is controlled by factors like precipitation and soil moisture. To achieve this, we used a Bayesian Auto-Regressive Distributed Lag (BARDL) modelling approach which enabled us to factor in lagged information from Precipitation and Soil moisture levels into our VCI forecast model. The results showed a ∼2-week gain in the forecast range compared to the univariate AR model used as a baseline. The R2 scores for the Bayesian ARDL model were 0.94, 0.85 and 0.74, compared to the AR model's R2 of 0.88, 0.77 and 0.65 for 6, 8 and 10 weeks lead time respectively.

A Dynamic Hierarchical Bayesian Approach for Forecasting Vegetation Condition

“…Relating the overall forecast skill assessments from this work to previous works, the HBM showed an approximately one week increase in the forecast range compared to the results from the BARDL method used in Salakpi et al (2021). On average, the HBM also exhibited an approximately 2-weeks increase in forecast range, compared to the auto-regression method used in Salakpi et al (2021) and Barrett et al (2020). Furthermore, using the HBM also enabled the simultaneous forecast of VCI3M for different AEZs and land covers which we could not do with the methods used in (Salakpi et al, 2021) and (Barrett et al, 2020).…”

“…Relating the overall forecast skill assessments from this work to previous works, the HBM showed an approximately one week increase in the forecast range compared to the results from the BARDL method used in Salakpi et al (2021). On average, the HBM also exhibited an approximately 2-weeks increase in forecast range, compared to the auto-regression method used in Salakpi et al (2021) and Barrett et al (2020).…”

“…On average, the HBM also exhibited an approximately 2-weeks increase in forecast range, compared to the auto-regression method used in Salakpi et al (2021) and Barrett et al (2020). Furthermore, using the HBM also enabled the simultaneous forecast of VCI3M for different AEZs and land covers which we could not do with the methods used in (Salakpi et al, 2021) and (Barrett et al, 2020). Finally, despite the improvement seen with the HBM, the BARDL models also proved to be useful at predicting drought events at the set threshold as demonstrated by (Salakpi et al, 2021).…”

“…Aside the AEZ boundary shapefiles, all the dataset used in this proof-of-concept is the same data used and described in a previous study here: (Salakpi et al, 2021).…”

“…In a previous study (Salakpi et al, 2021), we used a Bayesian regression method to model the relationship between biophysical drivers and their effect on forecasting vegetation conditions. The approach was based on the classical 'No-pooling' method (See figure 1), where we fitted separate regression models to data extracted from their respective regions.…”