Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, 91109, USA.
Clark University Graduate School of Geography, Worcester, Massachusetts, 01610, USA.
Ecol Appl. 2019 Mar;29(2):e01834. doi: 10.1002/eap.1834. Epub 2019 Jan 28.
This research investigates ecological responses to drought by developing a conceptual framework of vegetation response and investigating how multiple measures of drought can improve regional drought monitoring. We apply this approach to a case study of a recent drought in Guanacaste, Costa Rica. First, we assess drought severity with the Standard Precipitation Index (SPI) based on a 64-yr precipitation record derived from a combination of Global Precipitation Climatology Center data and satellite observations from Tropical Rainfall Measuring Mission and Global Precipitation Measurement. Then, we examine spatial patterns of precipitation, vegetation greenness, evapotranspiration (ET), potential evapotranspiration (PET), and evaporative stress index (ESI) during the drought years of 2013, 2014, and 2015 relative to a baseline period (2002-2012). We compute wet season (May-October) anomalies for precipitation at 0.25° spatial resolution, normalized difference vegetation index (NDVI) at 30-m spatial resolution, and ET, PET and ESI derived with the Priestley-Taylor Jet Propulsion Laboratory (PT-JPL) model at 1-km spatial resolution. We assess patterns of landscape response across years and land cover types including three kinds of forest (deciduous, old growth, and secondary), grassland, and cropland. Results show that rainfall in Guanacaste reached an all-time low in 2015 over a 64-yr record (wet season SPI = -3.46), resulting in NDVI declines. However, ET and ESI did not show significant anomalies relative to a baseline, drought-free period. Forests in the region exhibited lower water stress compared to grasslands and had smaller declines, and even some increases, in NDVI and ET during the drought period. This work highlights the value of using multiple measures to assess ecosystem responses to drought. It also suggests that agricultural land management has an opportunity to integrate these findings by emulating some of the characteristics of drought-resilient ecosystems in managed systems.
本研究通过建立植被响应的概念框架并探讨多种干旱指标如何提高区域干旱监测水平，来研究生态系统对干旱的响应。我们将此方法应用于哥斯达黎加瓜纳卡斯特地区最近一次干旱的案例研究。首先，我们使用基于全球降水气候中心数据和热带降雨测量任务以及全球降水测量卫星观测数据组合得出的 64 年降水记录的标准降水指数（SPI）评估干旱严重程度。然后，我们在干旱年份（2013 年、2014 年和 2015 年）与基线时期（2002-2012 年）相对比，检查降水、植被绿色度、蒸散（ET）、潜在蒸散（PET）和蒸发胁迫指数（ESI）的空间格局。我们以 0.25°的空间分辨率计算降水的湿季（5 月至 10 月）异常、30m 空间分辨率的归一化差异植被指数（NDVI）以及利用 Priestley-Taylor Jet Propulsion Laboratory（PT-JPL）模型得出的 1km 空间分辨率的 ET、PET 和 ESI。我们评估了跨年和土地覆盖类型（包括三种森林[落叶林、原始林和次生林]、草原和农田）的景观响应模式。结果表明，瓜纳卡斯特地区 2015 年的降雨量达到了 64 年记录中的最低点（湿季 SPI=-3.46），导致 NDVI 下降。然而，与无干旱基线期相比，ET 和 ESI 并没有表现出显著的异常。与草原相比，该地区的森林表现出较低的水分胁迫，NDVI 和 ET 在干旱期间的降幅较小，甚至有些增加。这项工作强调了使用多种指标评估生态系统对干旱响应的价值。它还表明，农业土地管理有机会通过在管理系统中模拟一些耐旱生态系统的特征，来整合这些发现。
