Lakes are unique in the land surface due to their well-known anomalous intrinsic properties – unusually large heat capacity, stark albedo contrast with water’s phase change and sun’s position. They also exhibit a wide spectrum of extrinsic properties – related to their occurrence, distribution and abundance. That they are ubiquitously besprinkled over most land surfaces with such intrinsic (extrinsic) anomalous (spectrum of) properties makes them a low-hanging fruit to observe from space and tease out in-land climate change signatures from local to global scales. In this dissertation, I explore four aspects in our current understanding of lakes and their connection to climate change, and I show that: 1) Long-term lake changes are multi-directional in nature as a rule and not exception; 2) Lake evaporation calculations using global data can be improved by ~5% at seasonal scales and ~50% (i.e. 5-10X better) in the energy gap to turbulence scales (i.e. ~30 minutes), compared to 5 other state-of-the-art mass-transfer methods, by virtue of a century-old misunderstood body of work by Robert E. Horton that is based on kinetic theory of gasses; 3) Long-term trends can be separated from correlation noise up to 1-sigma better than current practice in terms of both statistical power and confidence by combining a portfolio of 16 methods from two families of trend detection tools from Econometrics (parametric) and Hydrology (non-parametric); 4) Building upon the results of 1-3, a quasi-analytical water albedo model, and derived lake and climate variables from many sources (including Google Earth Engine datasets) can help us characterize lake changes up to sub-daily and sub-meter (micro-topography) scale, under the assumption of regional hydro-climate homogeneity at 0.25 degree spatial resolution (an unavoidable caveat governed by rain gauge density) for millions of Arctic Boreal Zone (ABZ) lakes. Collectively, these results I demonstrate at continental scale spanning whole of Canada and Alaska, ~10 million lakes, carve a pathway for a high-fidelity understanding of local-to-global scale climate change signatures on ~100 million lakes, which are, due to their intrinsic and extrinsic characteristics, our best in-land sentinels of climate change.